Bahama prehistory: cultural adaptation to an island environment

Item Type text; Dissertation-Reproduction (electronic)

Authors Hoffman, Charles Andrew, 1929-

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 28/09/2021 02:53:21

Link to Item http://hdl.handle.net/10150/565140 BAHAMA PREHISTORY;

CULTURAL ADAPTATION TO AN

ISLAND ENVIRONMENT

by

Charles Andrew Hoffman, Jr0

Dissertation Submitted to the Faculty of the

DEPARTMENT OF ANTHROPOLOGY

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1967 THE UNIVERSITY OF ARIZONA

GRADUATE COLLEGE

I hereby recommend that this dissertation prepared under my

direction by Charles Andrew Hoffman* Jr0______

entitled Bahama Prehistorys Cultural Adaptation to"an Island Environment ______'

be accepted as fulfilling the dissertation requirement of the

degree of Doctor of Philosophy______

JUULSU3& ■dU^yfi/oL Dissertation Director Date

After inspection of the dissertation, the following members

of the Final Examination Committee concur in its approval and

recommend its acceptance:*

8 / 3/ 6.7

f 1 ) iLHlos**. t L _ j’/V / £ 7

*This approval and acceptance is contingent on the candidate's adequate performance and defense of this dissertation at the final oral examination0 The inclusion of this sheet bound into the library copy of the dissertation is evidence of satisfactory performance at the final examination. STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED PREFACE

The possibility of conducting archaeological

investigations on San Salvador was first suggested to me

by Dr. Irving Rouse of Yale University. Dr. Rouse made arrangements several years ago for me to carry out excavations on that island. Although this project was

postponed until 1965, he has continued to generously give advice and support of my efforts in many parts of the

Caribbean area.

The field work and laboratory analysis for the

project were supported by a grant from the National

Science Foundation (05-7,58) to the University of Arizona

for Doctoral Dissertation Research in Anthropology.

Four institutions cooperating in the study were:

the University of Arizona, Florida Atlantic University, the

New World Museum on San Salvador Island, and Inter American

University of Puerto Rico. All research was carried out

under the guiding hand of my dissertation committees Dr.

Raymond H. Thompson, chairman, Dr, T . Patrick Culbert, and

Dr. William A. Longacre. Dr. Thompson has been a constant

source of moral and intellectual guidance, not only in

these last days of preparation of the results of the

ill iv findings on San Salvador, but throughout my graduate study program at the University of Arizona„

At the time of the field work I was an instructor at Florida Atlantic University and had the opportunity of being associated with Dr. William H, Sears, chairman of the Department of Anthropology at that institution. Dr.

Sears, who was probably the individual most responsible for launching me into a career as an anthropologist, pro­ vided much in the way of;support and advice and all too frequently in patience as virtually all the material processing was carried out in his department0 s laboratory.

We would never had been able to conduct this ! study on San Salvador if it had,not been for the kind consent of Mrs. Ruth G. Wolper, director of the New World

Museum, located on that islandMrs. Wolper. gave permission to dig on her property and permitted me to use resources of her Museum. Finally, I am pleased to acknowledge a small grant from Inter American University during the fall of 1966 that assisted in defraying some laboratory expenses.

Analysis of fauhal material was done by Dr.

Elizabeth S , Wing of the Center for Zooarchaeological

Research at the Florida State Museum. Dr. Wing or I will provide a copy of her synopsis on request. She is presently preparing a more detailed study for publication. V

Analysis of the flora was carried out by fflr. James Knowles of the Division of Plant Industries of the State of

Florida. His report is also available by directing

. A requests to him. Classification of rocks and minerals was by Fir. Richard Sense, graduate student in the Department I of Anthropology, University of Arizona.

All drawings were prepared by Miss Valerie

Jackson, graduate student in the Department of Anthropology, University of Arizona 5 photographs were taken by Miss

Helga Teiwes, Arizona State Museum photographer

I would also like to express my appreciation for

' ' ; the encouragement of Dr. Kenneth Bennett, Mr. Walter

Birkby, Mr. Jonathan Cell, Dr. William J. Kennedy, and

Mr. James V , Sciscenti. Their comments and criticisms have been most stimulating.

I suppose that all major endeavors, and a doctoral dissertation certainly seems to qualify as one, cost something besides money and material goods. And I also suspect that mine is not a unique case where those who paid the most were the family^of the candidate. And so, the greatest vote of appreciation is due to my wife, Eunice, and our children, for their perserverance and endurance. TABLE OF CONTENTS

. Page

LIST OF ILLUSTRATIONS viii

LIST OF TABLES ix

A BSTRACT @ * * » » * » » * » ® @ ® * o ® x

CHAPTER 1-INTRODUCTION I

San Salvador o « ® ® « * © © © © « ©• © « © © S Economy of San Salvador 9 Sketch of Prehistory of the Caribbean Islands , 12 Archaeology of San Salvador Prior to 1965 « , . 19 Some Ethnographic Observations ..,,©©.. 22

CHAPTER 2-THE 1965 EXCAVATION 28

The Palmetto Grove Site . © . © 29 The Excavations © © © © © © © © © © © © © © © © 32 C a t a 1 o g i n g o©@*©©©©©©©©©©©©©© ^ 1

CHAPTER 3-CERAMICS 43

Ceramic Attributes © . . .,.©.©..... 50 Palmetto Ware© » © ©© © © © * © » © © © © © © 56 ''Trade” Wares . 57 Comparison With Other West Indian Ceramics » , 59 Comparison With Mainland United States Ceramics 69

CHAPTER 4-OTHER ARTIFACTS „ 71

Stone © . . . © © © © © © © . © . © © © © © © © 71 Water Trough © © * © © © © . ©^ . © . © © © 72 Limestone Slabs With Shallow Grinding Depressions ...... 72 Grinding Stones . * © . © © © © © © © © © © 76 Hammerstone-Planer ...... 77 Beads© © © © # © © © © © © © * © © © © © . 78 Problematical Stone Objects ...... 78 8one .©.©.©.©a©©©...©*©.©* 82

vi vii

Page

Shell ...... 82 Shell Scrapers ...... 84 Shell Graving Tools „ ...... 88 Shell Drills ...... 93 Shell Weights . . . . . » ...... 94 Shaft Scrapers ...... 97 Strombus Shed 1 Gouges ...... 99 Strombus Hammers ...... 104 .Charonia Vessels ...... 105 Horn ...... 105 Strombus Ladle ...... 105 Cassis Chopper ...... 108 Shell Beads ...... 108 Oliva Tinklers ...... 110 Food Shells ...... Ill Miscellaneous ...... 112 Coral ...... H 4 Coral Hone ...... 114 Coral Corn Shockers ...... 115 Coral Rasps ...... 115 Coral Files . . . * ...... * • ... • 116 Problematical Coral Objects ...... 116

CHAPTER 5-SUMMARY AND CONCLUSIONS ...... 117

Conclusions ...... 122

REFERENCES 127 LIST OF ILLUSTRATIONS

Figure Page

1. Map of Bahama Islands ...... 2

2. San Salvador» Bahama Islands ...... 11

3. Extent of Excavations at the Palmetto Grove Site ...... 33

4 . Profile, East Wall, Trench E , Sections 10-12 . 35

5 . Stone Water Trough From Palmetto Grove Site . . 37

6. Shells From Conch "Pile." ...... 38

7. Extent of Heavy Fire and Charred Rock Area . . 40

8 . Ceramic Attributes on Pottery From Palmetto Grove ...... 52

9. Limestone Water Trough ...... 73

10. Stone Tools ...... 75

11. Shell, Bone and Stone Ornaments ...... 79

12. Shell Scrapers ...... 85

13. Shell Gravers and Drills ...... 89

14. Strombus Shell Weights ...... 95

15. Shell Shaft Scrapers ...... 98

16. Strombus Shell Gouges ...... 101 17. Miscellaneous Shell Implements . . . . , . . . 106

18. Cassis Horn ...... 107

19. Dove Shell Artifacts ...... 113

viii LIST OF TABLES

Tab le Page

1. Average Bahamas Temperature . . „ 7 2 . Average Monthly Rainfall, Nassau ...... , 8 3. Average Monthly Rainfall, Bahamas 8

4. Vertical Distribution of Palmetto Ware Pottery ...... 44

5. Vertical Distribution of Pottery Other Than Palmetto Ware ...... 45

6. Vertical Distribution of Shell Scrapers .... 87

7. Vertical Distribution of Graving Tools .... 90

8. Vertical Distribution of Shell Drills Cut from Strombus oioas ...... 91 9. Vertical Distribution of Shell Gouges n cj T i p s . 10 2

10. Vertical Distribution of Shell Beads and Tinklers . . . . . , . , ...... 109

11, Suggested Chronology for Prehistoric Caribbean Area ...... 123

ix ABSTRACT

Archaeological investigations on San Salvador

Island in the Bahama Islands are discussed in an attempt to better understand the relationships between culture and environment. After a brief introduction to the physical conditions of the island, its climate, general topography and economic status, the known prehistoric culture setting is sketched.

The prehistory of San Salvador is tied to the

West Indian islands rather than to the North American con­ tinent, Its pre-Columbian inhabitants were apparently a part of an Arawakan-speaking group that inhabited all the islands and parts of northern South America, Previous archaeological research is summarized. Some ethnographic observations made by Columbus on his arrival in the New

World for the first time provide the best known description of the peoples living in the Bahamas, The possible relationship between the Indians Columbus met and the people who lived at the site that was excavated is discussed.

Approximately 300 square meters wire excavated to a depth of 40 cm. at the Palmetto Grove site on the xi northwestern side of the island. Several kinds of pottery were recovered, including a shell-tempered red ware unique to the Bahamas, and several sand-tempered or grit-tempered, wares that resemble pottery from the Greater Antilles.

The shell-tempered pottery is tentatively named Palmetto

Ware. Shell tools, rather than potsherds, constitute the bulk of the collection. Because of an absence of hard stone on the island, shells, particularly the large conch,

Strombus qiqas. were used as a substitute. Several shell artifact types are introduced and special attention is called to the place of shell in the ecological adaptation of the Indians of an island such as San Salvador. Fish bones account for 99 per cent of the faunal material re­ covered; over 50 per cent of the fish are algae-eaters, mostly parrotfishes.

The few design elements on pottery from the

Palmetto Grove suggest that it could have been coeval with the Ostionoid horizon in the West Indies. This horizon style is normally considered to be ancestral to the ffleillac and Carrier styles, both of which heretofore had been reported as the dominant styles of the Bahamas. It is tentatively concluded that Palmetto Grove occupation was part of late Ostiones and ancestral to Meillac. On the basis of radiocarbon dating of the Ostionoid Horizon in the Greater Antilles, the site was probably occupied From

A.D. 750 to 1200. A combination oF marine and agricultural resources was probably the basis oF the prehistoric adaptation to the island environment. The sea not only

supplied Food, but also raw materials in the Form oF shells, corals, and Fish bone. CHAPTER i

INTRODUCTION

During the summer of 1965 archaeological in^> vesfcigations were undertaken on the island of San

Salvador in the Bahama Islands, The purpose was to learn something about the relationship between pre<= historic cultures and their environments, in general, and specifically to add to our knowledge of the. cultural prehistory of the Caribbean area. To this end the Palmetto Grove site on the northwestern coast of San Salvador was excavated. The materials recovered provide the basis for the interpretations presented in this study.

The Bahama Islands consist of an archipelago of islands, cays, rocks, and reefs lying off the southeast coast of Florida, and the north coasts of Cuba and

Hispaniola, It stretches roughly 700 miles in a northwest southeast direction between 20° 50* and 27° 51 * North

Latitude and between 71° 00° and 80° 25° West Longitude

(Fig, 1),

Included in the Bahamas are some 29 islands, over

600 cays, and more than 2,000 rocks and reefs. The 80 GRAND BAHAMA I. 76 74 72 70 24

O MIAMI ^BIMINI I ELEUTHERA I.

NASSAU

CAT I.

ANDROS I. 22 SAN SALVADOR I.

CAICOS I. 20

QGt. TURK I.

100

miles HAITI figure 1. Wap of Bahama Islands. N) archipelago extends from Great Inagua on the south to as far north as the Matanilla Reef and Grand Bahama Island.

The closest island to the United States mainland is Bimini, which is about 50 miles due east of Miami.

Great Inagua lies 60 miles north of the Cuban coast and

80 miles northwest of Haiti.

San Salvador

The island with which this report is concerned was named San Salvador by Columbus on 12 October 1492. It was later re«»named Wat ling’s Island by the British, and several years ago reverted back to San Salvador. It was originally called Guanahani by the Lucayan Indians who occupied it when Columbus discovered America. San Salvador is located along the eastern flanks of the Bahamas, about 380 miles east=-southeast of Miami, about 215 miles north of Cuba at its closest point, and 300 miles from the island 6f

Hispaniola. The nearest islands of any size are Cat Island,

45 miles to the west=northwest and Rum Cay, 23 miles to the southwest.

The islands of the Bahamas have not been studied much by geologists or geographers interested in their physical appearance. Perhaps the most significant work is that by Shattuck (1905). Numerous small papers have been published, particularly by the Institute of Marine Science in Miami, Florida; these usually bear on specific points. Basically $ the islands consist of Tertiary lime­ stone modified by vigorous coralline action, the actual sequence of which is open to debate. The rock is re­ ferred to as "new rock" or. "beach rock" because it is actually growing before the viewer's eyes. Growth is attributed to deposition of calcium carbonates from the sea water as waves pound up on the beach. New rock is also composed of many objects, such as large shells, tin cans and bottles that become imbedded in the rock as it forms. Paradoxically, as the sea deposits in one place it may erode in another, producing the razor-sharp sponge-cake rock of which a good part of the coastline of the Bahamas is composed.

There are no indications that the Bahamas were ever connected to the North American mainland. All fossils (Shattuck 1905) from the islands show clear affinities with Cuban and Central American types, all date from the Pleistocene, and all represent living .

The islands of the Bahamas are almost completely flat. The highest point of land is reported to be on Cat

Island, and is slightly over 400 feet above sea level.

For this reason, and because of the porosity of the lime­ stone, there are no rivers or streams, except occasionally on Andros. There is little fresh water on some islands and none on most„ San Salvador is one of the few having fresh water lakes.

On San Salvador the highest land is 141 feet in elevation, although the inland topography is generally rocky and rugged. There are numerous inland lakes where the water is brackish, but potable.

Seven surface soil types are reported in the

Bahamas, including three loam, three marl and one coral.

The three loam types are, of course, the most fertile and are fairly well suited to cultivation. Bahama Black Loam, the principal soil, occurs on most islands. It may occupy as much as three^fourths of the soil area of an island, although it must be remembered that surface soils are usually extremely thin except where lodged in cracks and crevices of the honeycombed limestone. There is another loam referred to as Pineapple Loam that is easily recog=> nized by its bright orange-red color, It is used for crops on San Salvador and especially on Eleuthera. It was apparently the soil selected for pottery making by some of the prehistoric Indians. For many years pineapples grown in this soil were a major crop in the Bahamas, hence the name.

The numerous limestone caves in the islands usually contain a soil deposit rich in- minerals known as cave<=earth, or guano. The islanders have removed most of this for use as a fertilizer.

The climate of San Salvador is typical of that for many of the Bahamas (Table l). Although the summer months are humid, the temperature rarely rises above 90° F „

Because of the humidity, few non-natives can endure the heat under outdoor working conditions. The mean yearly temperature is 78° F, The mean maximum of the warmest months (July-August) is 88° F ; the mean minimum of the coolest month (December) is 63° F .

There is a definite rainy season (Tables 2,3) from

August through November, although the annual total averages only 34 inches. There is a lesser rainy season in the spring. Prevailing winds are out of the east at about ten miles per hour and they apparently blow constantly.

The islands are in the path of the hurricanes, but for some reason they have rarely been severely hit. In

September, 1965, when a hurricane remained stationary over

Nassau for 24 hours, surprisingly little damage was done.

During August hurricanes center along the eastern edge of the Bahamas (in the area of San Salvador), but by Septem­

ber they have shifted to the western edge. By October the

line seems to shift well east of the islands. The most destructive recorded hurricanes have hit in 1866, 1873, 7

Table 1. Average Bahamas Temperature,

Following is the temperature For an average year for the Bahama Islands, taken from the official meteorological tables of the Bahama Chamber of Commerce (1952),

Month______Mean Temp,_____ Minimum_____ Maximum R

January 72 67 77 10

F ebruary 69 65 74 9

March 73 69 79 10

April 78 75 ;■ 84 9

May 80 75 85 10

June 79 74 85 11

July 81 76 86 10

August 82 77 88 11

September 81 75 88 13

October 79 73 85 12

November 75 68 82 14

December 70 63 77 14 8

Table 2« Average Monthly Rainfall, Nassau* The average monthly rainfall for a 40=-year period, at Nassau, New Providence Island, Bahamas (Yearbook of Agriculture 1941).

Jan Feb Mar Apr May Jun Jul Auo Sep Oct Nov Dec

2.14 1.61 1.53 2,51 5,89 6.66 5.83 6.59 7.55 6.34 8.25 1.49

Table 3. Average Monthly Rainfall, Bahamas.

The average monthly rainfall for all the Bahama Islands (Yearbook of Agriculture 1941)

Jan Feb May Juf Seo Oct Nov Dec

1.41 1.15 1.22 2.45 4,15 3.58 2.69 2.70 4.63 5.51 3.20 1.16 1888„ 1908, 1926 and 1945, The two worst were in 1866 and 1945,

San Salvador (Fig, 2 ) is approximately 12 miles long, north to south, and 6 miles wide. Its 60=square= mile area contains a brackish lake ten miles in length.

The island is almost completely surrounded by live coral reefs. There are few major channels through the reefs. Inside the reefs, in many places, there is quiet water and often gently sloping, sandy beaches. For the family living on the island the reefs provide a cdn= siderable amount of food, some protection from the ocean, and a sourbe of technological materials, but to the sailing visitor they were (and still are) frequently disastrous.

Economy of San Salvador

The present-day economy of San Salvador is based on two quite different factors. Generally, the people live on a combination of fishing and subsistence farming.

But this has more recently been modified by the cash economy of the United States government installations, and by tourism. Indeed, it was my observation that many of the younger generation did not know how to farm and if the United States removed all its facilities there, it ; 10 would be exceedingly difficult for the young people to carry on without moving off the island.

On San Salvador there are fewer than 1,000

Bahamans, of whom only one is of European descent (the

Catholic priest). About 100 of the native people are presently employed by the United States, but at one time this figure was well over 400. Between 1957 and 1954 there were on the island a guided missile tracking sta­ tion, a Navy facility, a Coast Guard station, and a Navy

Construction Batallion. Today the largest, the tracking station, has all but closed down, and the Sea Bees are gone„ Tourism, however, has not yet come to San

Salvador.

More to the point of my research, outside of wages, the people live by means of fishing and farming. Whether the soil is the best or the waters considered to be highly productive is not known, but a study of the feasibility of utilizing the natural resources should be of considerable value„ The soil seems thin, poor, and overworked, yet I had no trouble purchasing maize, guinea corn, pigeon peas, rice, and cassava during my stay. In fact, virtually all the vegetables that I ate while on the island were locally grown, as were many of the fruits.

It should be stressed that it is unlikely that any particular individual practices fishing or. subsistence 11

.CUT ROCK SITE

PALMETTO GROVE

BLUFF SITE O

COCKBURNTOWN

CREEK SITE 24

50

miles

figure 2. San Salvador Island, Bahamas 12 farming, that is, one to the exclusion of the other. By far the most common method of support is to plant a small garden or assist in a larger, cooperative farm. fishing is far less structured in that no particular person fishes to the exclusion of farming. It is interesting to note that many of the young men supplement cash income by fishing, but unless forced to, rarely by helping to farm.

In addition to government installations, cash in­ come may be obtained by working in small shops in the largest town (Cockburntown), or along the Queen’s Highway, which encircles San Salvador. Further, some men and a few women are on the local government payroll

Sketch of Prehistory

of the Caribbean Islands

The pre-Columbian relationship of the Bahaman

Islands to the United States and the remainder of the West

Indies has been bantered back and forth by experts and non­ experts for many years (Gower 1927, Holmes 1894, Loven

1935, Rouse 1949, Steward 1948). The general consensus has been that there was little if any cultural connection between the United States mainland and any of the Carib­ bean islands, including the Bahamas. But the fact remains that the Indians of the Bahamas did have a word for the land now known as Florida and for the natives of that 13

land (Major 1870)„ so at least there was an awareness of each other's presence„ if not any significant inter-mixing.

The issue is not dead„ however» and we shall probably continue to see papers on comparisons between the two areas. The burden of proof now seems to be more

upon the person interested in establishing a connection,

A tremendous amount of research has been undertaken in

the past five or six years by many students of the Carib- i bean, and prior to that by one man in particular (Irving

Rouse of Yale University), The first real evidence of

the connection between West Indian prehistoric cultures and any mainland came only after archaeological investiga­

tions by Rouse in the Greater Antilles, Trinidad, and

northern South America, Since then, work in the Lesser

Antilles has tended to substantiate Rouse's thesis that

the island Indians came from the South American continent.

There appear to have been at least three cultural

groups occupying the West Indian islands when Columbus arrived? the Ciboney, Arawak, and Carib, Brief mention

is essential to an understanding of the place the Bahamas

played in the overall picture.

The earliest occupants of the West Indian islands were apparently marginal groups called collectively the

Ciboney, They may have lived throughout the Caribbean at one time, but their exact distribution, where they came 14 from, or even how they entered the Antilles, is not known.

There is no record of them in the Bahamas despite

Cranberry8s exhaustive search of the literature (1955) and reconnaissance on Bimini (1957) and Grand Bahama

(Bullen 1963), or my own investigations on San Salvador or

Cat Island,

Ciboney remains are generally found in shell heaps along, the coasts, or in caves and rock shelters. They had no pottery and presumably no agriculture and according to ethnographic reports could not converse with other

Indians near by. They existed primarily by hunting, fish­ ing and gathering. By 1492 they had been driven to the western end of Cuba and tips of peninsulas on Cuba and

Hispaniola, For further information see Bullen (1963),

Cranberry (1957), Harrington (1921), Hatt (1924), Loven

(1935), Osgood (1942), and Rouse (1948, 1960, 1962, 1964).

Around the time of Christ a second, more sophisti­ cated, group entered the West Indies from the mainland of

South America, These were probably the Arawak, a peaceful, agricultural people^ tile can trace their movements through their pottery as they drifted northward from site to site, first in the Orinoco Valley, then along the Venezuelan coast, and finally in the Lesser Antilles, The time sequence for these sites is well marked by radiocarbon dating (Rouse 1963). It is assumed these people had some form of agriculture because the present-day Arawaks on the 15

mainland have the same (Roth 1924)„ and because of the

appearance of griddles in their refuse. Griddles have

been ethnographically associated with manioc agriculture

and the association is carried over into the archaeology.

Early ceramic sites in the Lesser Antilles have been

reported by Barton (1953), Bullen (1964, 1965), Haag

(1965), Loven (1935), McKusick (i960), and Rouse (1948,

1962, 1964),

As a result of the recent activity in the Lesser

Antilles, Caribbean archaeologists are now beginning to

see a break in the ceramic continuity of that string of

islands. This break is revealed by a slight shift in

ceramic design. It is not a complete break as there is

still a continuity of this design.

What was responsible for this shift is not exactly

known and the subject is one of considerable debate.

There are several alternate explanations. Such a shift

may have been the result of a second movement of Arawaks

from the mainland, bringing with them new ideas to add to

'the old. It could also have been produced by the movement

of ideas alone. As a third alternative, it may have been

the introduction of the Caribs, who were not supposed to

have arrived until much later. In the latter case the

Caribs would have subdued the Arawaks and taken the Arawak=

an women as wives, a situation which could produce a noticeable departure from, but not a complete break with the previous ceramic styling. The time period for this shift was roughly A .D.

600. For further information on the archaeology of the subsequent period see Barton (i960), Bullen (1964, 1965),

Haag (1965), Hoffman (1963), fflcKusick (i960), and

Rouse (1962, 1964).

When the Arawaks first entered the islands they pushed on as far as Puerto Rico, the Ciboney apparently giving way before them. The extension of the Lesser

Antilles culture at this period of time, referred to as the Saladoid Series (Rouse 1964), has been radiocarbon dated as early as A.D. 120 t 80, at the Hacienda Grande site near San Juan, Puerto Rico (Rouse 1963, 1964).

At roughly the same time that we see the A.D. 600 shift in ceramics in the Lesser Antilles, the Arawaks in the Greater Antilles apparently pushed on to the islands of Hispaniola and Jamaica, and shortly after to eastern and central Cuba. Finally they moved into the Bahama

Islands. By about A.D. 900 the Arawaks had occupied all but the western tip of Cuba. This movement is documented by the spread of ceramic stylings and radiocarbon dating of these stylings (Rouse 1963, 1964).

The third major culture to enter the West Indies was the Carib, a warlike, cannibalistic group. Although they did practice agriculture and were really quite similar 17

to the ArawakSj, culturally and physically, they were at

the time more aggressive and continuously on the move„

When they entered the West Indies is presently not known.

It has been assumed they had only been in the islands

about 100 years when Columbus arrived there on his

second voyage. This latter is according to an ethnographic

account (Major 1870) of the time. Because we do see a

change in the archaeological material between A.D. 600

and 800, it may have been the Caribs who were responsible for that change and that they arrived in the Antilles much earlier than heretofore believed. The Caribs killed and

ate the Arawakan men (a practice reported for both Caribs and Arawaks in more recent times in the adjacent areas of

South America /Roth 1924%).

To summarize the above % the islands of the Carib=»

bean were inhabited by at least three groups of Indians.

The earliest known was the Ciboney, a non-ceramic culture.

The Ciboney were apparently distributed throughout the

Caribbean. Columbus found remnants of this group in re~

mote places on Hispaniola and Cuba, where they existed

primarily by hunting and fishing. Their remains are

characteristically along the coasts and in caves. They

had no pottery and presumably no agriculture. Perhaps as

early as A.D. 1 a second, more sophisticated group entered

the West Indies from the mainland of South America. These

were the Arawak, a sedentary people. Their presence is 18

marked by the appearance of pottery and agriculture.

Sometime between A.D, 600 and 800 we see minor changes in the Arawak ceramics. This change may be attributed

to influences out of South America, in the form of a) idea diffusion, b ) more Arawaks, or c) the Caribs,

The Carib, a warlike, aggressive Indian who practiced

cannibalism, was thought to have entered the West Indies about 100 years before Columbus arrived there. They

swept northward from the lowlands of South America,

moving from island to island, killing off the male

Arawaks and taking their women and children as slaves.

By 1492 they were making stabbing advances against the

coasts of Puerto Rico and Hispaniola,

One of these three groups of Indians lived on

San Salvador and deposited the material collected and

described in this report. Presumably they were Arawaks,

chiefly because the presence of pottery would rule out ■

the Ciboney, and because it is doubtful that the Caribs

had gotten that far north. In addition Columbus (Major

1870) reported finding a peaceful people on San Salvador,

who apparently had no knowledge of weapons, Columbus

also noted the similarity between the language of the

Lucayans and that of the Arawaks of the Greater Antilles.

In fact he kidnapped six Indians from San Salvador to

serve as interpreters as he continued on his journey

through the West Indies. 19

Archaeology of San Salvador

Prior to 1965

Little anthropological work has been Conducted

in the Bahama Islands, In his exhaustive treatment of the archaeology of the Bahamas Cranberry (1956: 128) lists

only eight surveys. Since then there have been three studies, including the one upon which this report is

based. Cranberry visited Grand Bahama (Sullen 1963),

Goggin excavated several sites on San Salvador in 19160

(unpublished), and I conducted a survey on Cat Island in 1965=66,

The earliest report of a specimen found on San

Salvador dates from 1828, A wooden duho, or ceremonial seat, was supposed to have been collected then, and is reported (Cranberry 1955) to be in the collections of the

Academy of Natural Sciences in Philadlphia,

Krieger (1937) mentions finding pottery, wood, shell, and stone in caves on San Salvador, but no listing of sites is given. The pottery sound similar to that re= ported on here as the major ware found at the Palmetto

Grove site on northern San Salvador, Rainey spent five days on the island looking for sites. He located four caves which produced cultural material. At the head of the central lake on the island (apparently not far from

Palmetto Grove) was found a nearly complete Meillac bowl. 20

Rainey also found ffleillac pottery in caves on the south central part of the island. Cranberry (1955) com= ments that they are thin, with a buffered surfacep typical of Great Abaco and San Salvador Pleillac sherds. In one cave were found 2 decorated ffleillac, 24 Carrier sherds and

5 fabric-impressed body sherds. I would assume from this

(Cranberry 1955s 149) that the fabrie®impressed sherds were of a different type, perhaps more like the fabric® or basketry®impressed, she11®tempered red ware found at

Palmetto Grove, and to which Krieger had referred.

Goggin excavated at the Palmetto Grove site in

1960. He dug six test squares and retrieved mostly the thick reddlsh-paste shell®tempered pottery mentioned above.

He also dug at least three other sites (fig. 2)s the Ward site located very close to Palmetto Grove but a few meters farther inland, the Bluff site, south of Long Bay on the west coast, and the Cut Rock site, out the northeast tip of the island.

Meanwhile, during the past ten years, the staff of the New World Museum, Mrs. Ruth G. Wolper and members of her family, have been surface collecting and excavating on

San Salvador. The bulk of the pottery recovered is the same thick red ware, although decorated pieces on display in the

New World Museum combine both Meillac and Carrier traits„

That is, modes normally found in only one, but not both, of these styles, are found on the same red ware vessel, Mrs. 21

Wolper located a large site on the southeast side of the island about a quarter«=roile from the beach„ at the head of

Pigeon Creek (Fig, 2), In addition she has excavated at the Palmetto Grove and nearby sites»

Carrier and Meillac apparently were the dominant styles in the Bahamas with the unnamed red ware possibly representing a third. Both ceramic and nonceramie specimens recovered from Bahaman sites show marked similarities to artifacts from north Haiti as defined by House (1941s

54=91, 113=40, Pis. 7=24, 27=34). Their distribution appears patterned, the Meillac being in the north and the

Carrier to the south with the middle overlapping. This would make sense as the Meillac was the forerunner of

Carrier in Haiti. Carrier probably did not have enough time to work its way to the north end of the Bahaman archipelago.

Of course, the movement of ideas reflects the degree of contact, in this case between islands. As is discussed below in greater detail, we can conclude from our knowledge of the fine navigation ability of the

Lucayans that it should not have taken very long for the later Carrier to permeate all of the Bahamas, With this inferred evidence we can postulate that the movement of

Carrier traits was arrested by the arrival of the

Spaniards, and must date very close to 1492. 22

Some Ethnographic

Observations

Ule have little in the way of ethnographic data on the Lucayan Iridians of the Bahama Islands, They were a branch of the Taino Arawaks, who occupied the Greater

Antilles, How long they were in the Bahamas is not known and probably will not be until we obtain radiocarbon dates on their sites. The pottery found in the islands indicates that they were not there many centuries before Columbus discovered America,

One of the reasons for not having much information is the short period of time that the Indians lasted under

Spanish influence. It is estimated that by 1513, only 20 years after Contact, the entire population of the Bahamas had been exterminated by the Spaniards (Las Casas 1952).

It is not the purpose of this report to go into the unhappy story of the demise of the Lucayan Indians, For this the reader is directed to Las Casas, the Hakluyt Society publications, or fflorison (1942).

Some of Columbus" observations (fflorison 1942s

228=31) have a direct application to the archaeology of San

Salvador and the Bahamas, The following is translated by

Morison from Las Casas, who was supposedly quoting

Columbus" exact words. It refers to the first landing by

Columbus in the New World and a second landing on the following day, 13 October 1492. In order that we might win good friend­ ship, because I knew that they were a people who could better be freed and converted to our Holy Faith by love than by force, I gave to some of them red caps and to some glass beads, which they hung on their necks, and many other things of slight value, in which they took much pleasure; they remained so much our friends that it was a marvel; and later they came swim­ ming to the ship9s boats in which we were, and brought up parrots and cotton thread in skeins and darts and many other things, and we swopped them for other things that we gave them, such as little glass beads and hawks9 bells. Finally they swopped and gave everything they had, with good will; but it appeared to us that these people were very poor in everything. They go quite naked as their mothers bore them; and also the women, although I didn9t see more than one really young girl. All that I saw were young men, none of them more than 30 years old, very well made, of very handsome bodies and very good faces; the hair coarse almost as the hair of a horse's tail and short; the hair they wear over their eyebrows, except for a hank behind that they wear long and never cut. Some of them paint themselves black (and they are of the color of the Canary Islanders, neither black nor white), and some paint themselves white, anti others red, and others with what they have. Some paint their faces, others the whole body, othjgrs the eyes only, others only the nose. They bear no arms, nor know thereof; for I showed them swords and they grasped them by the blade and cut them­ selves through ignorance; they have no iron. Their darts are a kind of rod without iron, and some have at the end a fish's tooth and others, other things. They are generally fairly tall and good looking, well made. I saw some who had marks of wounds on their bodies, and made signs to them to ask what it was, and they showed me how people of other islands which are near came there and wished to capture them, and they defended themselves. And I believed and now believe that people do come here from the mainland to take them as slaves. They ought to be good servants and of good skill, for I see that they repeat very quickly all that is said to them; and I believe that they would 24

easily be made Christians9 because it seemed to me that they belong to no religion, I„ please Our Lord* will carry off six of them at my departure to Your Highnesses, so that they may learn to speak, I saw no beast of any kind except parrots in this island,

Saturday October 13a At daybreak there came to the beach many of these men, all young men as I have said, and all of good stature, very handsome people. Their hair is not kinky but loose and coarse like horsehair; and the whole forehead and head is very broad, more so than any other race that I have seen, and the eyes very handsome and not small, and themselves not at all black, but of the color of the Canary Islanders; nor should anything else be expected, because this is on the same latitude with the island of Ferro in the Canaries, Their legs are very straight, all in a line; and no belly, but very well built. They came to the ship in dugouts which are fashioned like a long boat from the bole of a tree, and all in one piece, and wonderfully made (considering the country), and so big that in some came 40 or 45 men, and others smaller, down to the size that held but a single man. They row with a thing like a baker's peel and go wonderfully, and if they capsize all begin to swim and right it and bail it out with calabashes that they carry. They brought skeins of spun cotton, and parrots and darts and other trifles that would be tedious to describe, and gave all for whatever was given to them.

In a letter that Columbus wrote to "the Sovereigns"

in Spain, some information about Spanish trade goods may

be found. Here he mentions (Morison 19421 231)s "I

forbade that they be given things so worthless as bits of

broken crockery and of green glass and lace-points, although when they could get them, they thought they had the best jewel in the world," What was given in

substitute is not recorded. 25

And sog for the archaeologist trying to recon» struct the culture that existed on this tiny island

Columbus6 meager account provides a wealth of information.

Apparently the Indians had no metals, except gold which he mentions later was hung "from a thing like a needle- case which they wear in the nose" (fflorison 1942s 237),

They had parrots, weaving, darts tipped with fish teeth, various kinds of paints, dugout canoes that would hold

40=45 men, paddles, calabashes and broad foreheads.

This last .was actually probably frontal deformation as indicated by Lucayan skeletal remains reported from the

Bahamas (Brooks 1888).

Of additional importance is the mention of trade goods, many of rather non-perishable materials. As we shall see, no trade gdods were found at the Balmetto

Grove site. Indeed, I know of no trade goods found on

San Salvador at all, despite the extensive digging by the

New World Museum on that island. If the sites being excavated were occupied when Columbus arrived, then one might expect to find glass beads, red caps, hawks0 bells, lace points, broken Spanish crockery, green glass, and such objects that the Spaniards simply lost during their trips to the island, or which were traded, but not reported. The bells are small round bronze bells used in falconry, the beads Venetian glass beads. The lace points were the metal tips of the laces then used instead 26 of buttons on men°s clothing„ On Long Island, the third to be visited by Columbus, and only 60 miles southwest of

San Salvador, the Spaniards left beads, lace points, and brass tambourine jingles. Whether or not the last mentioned were also left on San Salvador is not known.

Also from Long Island comes the first report of the New World hammock, A watering party visited several native huts at the extreme northern end of the island and reported (fflorison 1942s 245)s "their beds and furnishings were like nets of cotton,” Morison then quotes Las Casass

These are here in Hispaniola called hamacas, which are in the form of slings, not woven like nets with the threads going zigzag, but the lengthwise threads are so loose that you can insert the hand and fingers, and at a hand’s breadth more or less they are crossed with other close-woven threads like well made lace-trimmings.

The hammocks were about 5^ feet long with the two ends finished off in many loops of the same threads, in each of which were inserted some delicate threads of another substance stouter than cotton, like hemp; and each of these was about 6 feet long. Not like the "back yard hammocks" in the United States, these may be from 8 to 11 feet wide. They are slept in at an angle. Hammocks were adapted by seafarers the world over and even today are used by the Indians of Yucatan, Central and South

America, Columbus provides additional information from

ALong Island, While his crew was away after water he

observed the native houses. These dwellings he com­

pared (Morison 1942: 246) in shape to Moorish tents

"very high and with good chimneys," being a sort of

open cupola to let the smoke out. His men reported the

huts to be "very simple and clean inside." The people

kept pet dogs which the Spaniards said did not bark.

One man wore a big gold nose-plug which Columbus was

eager to possess because it seemed to have letters which

he suspected to be the inscription on a Japanese or

Chinese coin, but the man would not part with it. It

was from this village that Columbus6 men obtained CHAPTER 2

.THE 1965 EXCAVATION

The primary purpose of the excavation was to attempt to increase our knowledge of cultural adaptation to island environments through the analysis and interpre­ tation of spatial and temporal distributions of arti- factual, faunal, floral and other materials. If successful the project would serve as a guide for future studies of a similar nature, aiding in improving our methods and techniques. Meanwhile I had hoped to find out more about the people who were, presumably, the first to be seen by Columbus upon his arrival in the New World, and to help establish the position of this island in the Bahaman and West Indian cultural picture.

San Salvador presented a good opportunity to undertake a study of cultural adaptation. The very fact that it is an island places certain definitive limits on the environmental situation. It is small and relatively isolated, and the prehistoric cultures were simpler than those of the larger, more central West

Indian islands.

Many of the desired accomplishments were not attained. Because of a shorter field session than

28 29 originally planned» only one site was excavated„ The pottery, upon which I had planned to conduct design analysis turned out to be mostly plain undecorated ware.

In addition, distributional analysis of bones is hindered by the unusually high prevalence of parrotfish bones.

All1 but one or two bones are from fishes (99/6) and about

50% of these represent the parrotfishes.

The Palmetto Grove Site

The Palmetto Grove site is located at the north end of San Salvador on the property of Mrs. Ruth G.

Wolper, director of the New World Museum. It is by a small bay, on the north slope of a low ridge. The ground slopes gently to the east and south. Where this ridge projects into the sea, west of the site, is known as Rocky

Point (Fig. 2).

The Palmetto Grove site is recorded here as

SAL-3. It should be pointed out that in the New World

Museum it is listed as SAL-3, SAL-2, and S.S.2. When

Goggin excavated there in 1960, he referred to it (unpub­ lished notes) as SS-2. It is also called the "Orange

Grove" by the New World Museum and "Ruthie1s Orange

Patch" in Goggin * s notes at the University of Florida

Anthropology Laboratory. The designation SAL-3 and the name "Palmetto Grove" are used here at Mrs. Wolper’s request. 30

The shallow bay stretching northeast of Rocky

Point has no name to my knowledge. This bay apparently

was once considerably more indented and might at one

time have cut back into a small lake (Tig. 2), south

and southeast of the site. During the course of the

excavation two piles of conch (Strombus gioas) shells

were found well beneath the occupation zone. The conchs

were distributed about in a manner similar to the way

they are piled up on beaches today. The impression given

was that that spot had once been a beach which subsequent­

ly filled in with sand upon which the Indians then

settled. I discussed this with William G . Haag when we

visited the site in June, 1966. He suggested that it

was highly possible and that it is a natural tendency

for the curved 1ine of bays and coves to straighten out.

During recent years tremendous quantities of sand have

been moved during a single storm. It seems quite probable

that where the site is now was once part of the bay.

The site area may be described in terms of

quarters: north, eastsouth and west. Looking northward

from the center, the ground rises gently to the crest of

sand dunesg whence it drops quickly to.the sea, about

130-40 m„ away. At its highest point this crest is about

5 m. above sea level. To the east the land slopes down,

toward a natural spring where fresh water is often within 31 reach, and to the southeast to a small saline pond. To the northeast it rises beyond the spring to meet the ridge forming the far side of the bay. To the south a paved road crosses the edge of the site. Beyond the ; road the land drops gently to the pond l&ss than 100 m. away. To the west the ground rises and tops out at the crest of the limestone ridge forming the backbone of

Rocky Point. This highest point is slightly more than

7 m. above sea level. Offshore there is a reef composed mostly of algae but also a few living corals.

When the occupants of the site lived there, they were probably on the edge of a bay that penetrated farther inland, near a saline pond, but with access to fresh water. They had the various fauna associated with tidal rocks, and offshore, a reef teeming with an unusually high concentration of algae-eating parrotfishes. The

Indians settled in back of the dunes, probably for protection from the sea. As much of the land is being . farmed today, it is quite likely that it may have been suitable for some farming then. The soil is chiefly beach sand, no marl being observed at all. Some black loam may have been present. The Indians probably practiced swidden, or slash-and-burn agriculture, just as the present-day inhabitants of San Salvador do.

Fresh water was also found beneath the south end of the site. The crew dug through clean yellowish-white 32 sand to 130 cm, where they encountered soft limestone forming in place and good-tasting sweet water.

As discussed below, the occupation level was probably at about 28 cm*, below the present surface, making it about 1 m, above drinking water. This would be the water table exposed in the spring when the latter is cleaned out, as it is only a few meters from the east edge of the excavations. Whether or not these condi­ tions prevailed when the Palmetto Grove site was occupied is not known.

Prior to clearing, the site was covered with brush and a few trees. Much of the underbrush had grown since 1963 as the site area had been cultivated by hand in that year. In addition to the brush, corn stobbs, apparently from the 1963 planting, thatch palm, sabal palm, and a few orange trees were present. Only the last are located in the area excavated and would account for some disturbance of the site when they were set in. Such disturbance would! have been small, however, and in any case I left most of them in place and dug around them.

The Excavations

The site was excavated in 2 m, squares and in arbitrary 10 cm, levels. The resultant map of excava­ tion (Fig, 3) was designed to produce a grid-work that would be amenable to distributional analysis. All dirt

Figure 3„ Extent of Excavations at the Palmetto Grove Site.

The excavated area is indicated by hachured lines; the unexcavated is blank. Gray area in center approxi­ mates the outline of Goggin0 s 1960 trenches A and B . Trenches are indicated by letters. Trees in the center are small orange trees. 33

G r E D G

PALMETTO GROVE

2 m.

o Iron pipe

Figure 3. Extent of Excavations at the Palmetto Grove Site. 34 was. sifted through quarter=inch mesh hardware cloth.

Every effort was made toward careful measurement, and the research design was such that the more of the site excavated the better. No burials or evidence of burials, or architectural features were encountered and digging moved along at a rapid pace.

What portion of the occupation area was excavated is not known. At the end of the season approximately 300 square meters had been covered. Test pits were dug around the site; all proved sterile, A few sherds were found on the surface of a newly planted field on the eastern, opposite slope, I could not dig this field, but it had recently been plowed and if the site extended over that far it was not evident from the surface,

I usually dug to 40 cm. On a few occasions it was necessary to dig to 50 and even to 60 cm,, but this was because of slightly higher ground in one particular spot, perhaps because of overburden from nearby farming, or unremoved back dirt from Goggines excavation. The concen<= tration of evidence was in the 20=30 cm, level. Wall profiles (Fig, 4) indicate the living surface was between

25=30 cm. This was marked, by dark, sometimes black stains and of course the concentration of material, A large

"water trough" was found in the southeast part of the site. It was too heavy (Fig, 5) to be easily transportable. The bottom of the trough was resting at Ash, Charred Rocks | | Brownish-gray Sand

Occupation Horizon | | Light, Sterile Sand

Humus Unexcavated

Figure 4. Profile, East Wall, Trench E , Sections 10-12. 28 cm. It is assumed that this represents the living surface. Because the vertical distribution of artifacts tapers off quickly above and below this point, I am also assuming the site was occupied only once and for a short period of time. The chronological position is discussed in Chapter 6.

During the excavation two "piles" of conch shell were uncovered» one in the southwest corner of the site

(around Section H-6) and the other to the north in the area of Section H-12. One of these piles (Fig. 6) was completely removed and amounted to at least 28 whole conchs. A distinguishing feature, which may be seen in the photograph, is the tiny round hole in the apex used to remove the meat» This hole was apparently put there by puncturing the shell with the apex or horn of another conch shell. It took a rather small instrument to fit into the hole for cutting the muscle of the .

These piles of conchs are similar to those found along the coast lines of many West Indian islands. Usua the conch fisherman removes the meat as soon as he gets back to shore, so he won't be burdened by carrying both the meat and the heavy shell back to his camp area. I have seen several such piles on San Salvador.

The two piles found in the excavation are appar­ ently unrelated to the occupation of the site; they are completely be low the level at which pottery was being 37

Figure 5. Stone Water Trough From Palmetto Grove Site. 38

Figure 6. Shells From Conch "Pile." 39 recovered„ At about 65 cm. In Section H~12 several conch horns protruded upward. These horns projected up into a fire area and are obviously burned, but the bottoms of the shells are not. Below these, others were not fire-marked, seeming to confirm their unrelatedness to the time the fire was actually burning.

An extensive fire area (Fig, 7) in the northwest portion of the site presents something of a puzzle, It is long, roughly rectangular in outline, and resembles present day charcoal pits. As with charcoal pits, many small rocks were found, all fired. Very few charcoal flecks were recovered, but in one section (between H=ll and H-12) two fragments of logs were recovered,

The following description generally covers the profile (Fig, 4) at Section H-12, Ground level to 8 cm, was humic, 8 to 40 cm.— gray sand, 40-50 cm. white-gray sand, 50-60 cm.--black sand and fist-sized rocks (some of the rock fragments were saved). In other sections the rocks appeared slightly higher up, say about 40-50 cm., and the ground was literally paved with them. The levelness of the top is not discernible, but the change is clear and seems to be level, There is a tremendous quantity of darkened, fire-baked rocks in the dark level. This horizon is generally 10 cm. thick. Then it gives way to yellow- white sand such as is found beneath the rest of the site. 40

17

16

15

14

13

12

11

10

9

8

7

6

5

7. Extent of Heavy Fire and Charred Rock Area. The Fire bed rises slightly toward the north, but not more than a Few centimeters.

It has been mentioned that this Fire bed is reminiscent oF charcoal pits. It is also possible that it is or was a lime kiln, thus explaining the many small

Fist-sized chunks oF limestone. It would also explain the almost complete absence oF charcoal oF any size in the center oF the bed.

The solution, or at least elimination oF some alternatives, will be nearer at hafnd when Carbon-14 dates are obtained From the charcoal, or IF perhaps thermo- luminescence-dating can be determined For the Fired limestone,

Figure 3 shows the location oF the two trenches,

A and B, excavated by Goggin in I960, His work was con­ ducted in Five-Foot squares, and dug in arbitrary six- inch levels. Six squares were excavated, the bulk oF the material being recovered in the 6-12 inch level. This correlates well with my own work.

Cataloging

All material recovered is catalogued in the New

World Museum, Catalog numbers were assigned to each level oF each square. All objects retained From a par­ ticular level, within a square, received the catalog 42 number of that section. The first square or section of the first level of the site became NWM-1 because it con­ tained the first set of items ever to be catalogued in the New World Museum. A tabulation of the- catalog numbers and their respective levels is on file at the

Museum.

i Catalog numbers may occasionally appear in this report, either on specimens in photographs, or in the text. This is solely to assist in identifying the specimen to which reference is made in case of later analysis of the collections. CHAPTER 3

CERAMICS

All of the pottery recovered from the Palmetto

Grove site is in the form of sherds; no complete vessels were found. Several kinds of pottery may be distinguished

(Tables 4, 5), but 98.8. per cent is comprised of a single ware characterized by several recurring identifiable structural traits; undecorated surface, shell-tempering, thickness and softness and reddish paste. Some of it is basketry-impressed.

The remaining types may or may not be considered

"trade wares.” Two of these are tempered with quartz sand, a material that is foreign to the island. One of these two is buff, sandy-paste, thin, well-fired pottery with one sherd decorated with an incised sigmoid applique.

Another type is intermediate between the latter and the soft reddish paste pottery. It is red to reddish black, as hard as the sandy-paste pottery, well fired, thin, and contains a few large grains of quartz apparently as temper. A fourth kind of pottery is accounted for by only five sherds. It is distinctive in that it is thin- walled, black in color, shell-tempered, and slipped with the same red clay that makes up the ever-present plain ware.

43 44

Table 4. Vertical Distribution of Palmetto Ware Pottery,

This table includes all sherds recovered although only 11 sections were excavated to 50 cm., and 5 to 50 cm.

txcavation Levels in Cm, Sherds T otal 0-10 10-20 20-30 30-40 40=50 50-60

Plain 376 903 2996 655 60 5005

Horns or Rimpoints 6 8

Strip Lugs 3 3

Finger Indented 1 1

Griddle 7 25 111 153

T otal 383 928 3117 666 69 5170 45

Table 5. Vertical Distribution of Pottery, Other Than Palmetto Ware.

This table shows the stratigraphic relationships of the pottery wares ^ that have not yet been formally named. None were recovered below 40 cm.

Excavation Levels in Cm. Pottery Wares Total 0=10 10=20 20-30 30-40

Grit Tempered, Reddish Paste Plain 4 9 5 18 Lunar, Wedge-Shaped Lug 1 1 Incised 1 1 Incised Lug 1 1

Red Slipped, Black Core Plain 1 1 1 1 4 With Strip Lug 1 1

Sand Tempered, Gray-Black Paste - 3 5 1 9

Sand Tempered, Buff Paste 4 1 5

Sand Tempered, Brick Red Paste 2 1 3

Shell Tempered, Gray Paste 1 3 4

\ Palmetto-like Ware, Thin, - Brown Paste 3 1 4

Red Slipped, Hard, Gritty 1 1

Palmetto Ware "Chunks" 2 25 1 28

Total 2 21 45 12 80 46

Some of the Palmetto Ware sherds are basketry impressed. Because of the softness of the pottery many of these impressions are difficult to see, much less classify. However, there seem to be at least four dif­ ferent types: checkerwork, twilledwork, wickerwork and twined. Examples and discussion of these categories of basketry may be found in Mason (1904; 228; Figs, 3, 5,

6, 22), There seems to be no reason to regard these impressions as having been made with a fabric, as stated by Cranberry (1955, 1956), They were definitely made with either basketry or matting.

Many basketry-impressed sherds are flat and often are fragments of griddles. It may be that the impression was produced by pressing soft clay onto a mat. It appears as though a mat had been placed on the ground as a working surface and a lump of soft clay was then spread out over the mat. As the clay was pressed flat the impressions would be produced on the under surface. Thus the basketry impressions do not seem to have been a form of

( decoration, but simply a by-product in the manufacture of griddles or flat-bottomed vessels.

It must be added at this point that decorated

Palmetto Ware apparently does occur. It simply was not found in my excavations at the Palmetto Grove site. In the New World Museum collections there are fragments of

Palmetto Ware bowls with incised shoulders, 47 modeled-incised zoomorphic lugs, and whose shapes are readily discernible. Mrs. Wolper told me that some of these vessels are from the Creek Site on the opposite side of the island from Palmetto Grove, and one or two are from Palmetto Grove itself. So far, such cases of decorated Palmetto Ware are rare indeed and in any event we do not know the chronological or associative context of the sherds.

In the collections at the University of Florida

Anthropology Laboratory there are several decorated sherds, both Palmetto Ware and of the paste referred to as Sand Tempered, Gray-Black Paste (Table 5). These sherds are in the pottery Goggin excavated from San

Salvador in 1960. However, I do not believe they came from the Palmetto Grove site. All the sherds listed as being from the latter are plain, whether Palmetto Ware or otherwise. z'; ^

Goggin dug test pits at the Ward Site, about a half-mile south of Palmetto Grove and perhaps 200 meters inland from the west coast. The sherds from this site are also Palmetto Ware, but are generally thinner and harder and a greater percentage of them seem to come from jars.

In his field notes Goggin reports that he recovered one punctated rim sherd from the surface of the Ward Site.

There is one sherd in his collection, a gray sandy-paste ware, that fits this description. It is apparently part 48

of a typical Me11lac bowl» with the squarish punctations

scattered about the rim in neat rows and with a sharply

inturning shoulder or keel below the rim.

At another site Goggin excavated. Cut Rock, he

found two 11 broad-lined incised" shell-tempered rim sherds on the surface. There are two such sherds in the UFAL collections. They are similar to Palmetto Ware except in being a little thinner. The incising is apparently part of a series of oblique lines just below the rim. Again, this would seem to indicate a Meillac influence.

Because the dominant pottery is essentially a plain ware, it is not possible to classify it as easily as one might according to standard type classification procedures. This is a problem similar to the one that

Goggin had when he was confronted with his Olive Jar sherds (Goggin I960), except that he was fortunate in being a little more certain of the shapes of his vessels than I am of the ones from San Salvador, As he states it (Goggin 19608 24-25)s

As a classificatory device the type concept seems most practical when it serves to bring together a series of noncomparable modes, such as paste, decorative technique, design, vessel shape, etc. However, when a series of alternate comparable modes are dealt with, the type concept becomes awkward unless each one is used as a type diagnostic itself, as in the case of Uleeden Island Plain, Uleeden Island Incised, and Wakulla Check Stamped, where the alternate comparable modes of "plain," "incised," and 49

"check-stamped” are really all that dif­ ferentiate the three types.

What has been done in this paper was to set up three ceramic divisions based on temporal differences. This is a compromise chosen as the most consis­ tent way to deal with all the material. Some could be treated as types. but not all, so the system was designed, as all classificatory systems should be, to deal with the material at hand rather than to follow any present conceptual framework.

The bulk of the pottery from Palmetto Grove might be set up as a type, but there is little variation within each of the non-comparable modes. The desire is to avoid obscuring something of cultural or historical interest such as might be the case if all the alternative modes are lumped into one type.

To refer to the dominant pottery as a type, at this stage, would produce only one? Such-And-Such Plain.

To go beyond that point more sherds are needed and more sites should be represented. However, to ignore the various alternative modes could result in our not being able to discern temporal and distributional suggestions, as well as frustrating future archaeologists in their attempts at the solution to problems with which they may be concerned. Obviously, then, in this case modal analysis cannot be avoided.

The treatment here of the .ceramics from the

Palmetto Grove site is essentially along the lines developed by Rouse (1939? 25). However, to facilitate 50 discussion of the dominant form of pottery a purely arbitrary term is here being introduced to serve as a substitute for its description. This term is "Palmetto

Ware." Palmetto Ware should imply to the reader a consistent combination of non«=-comparable attributes which recur from sherd to sherd.

Ceramic Attributes

A list of the attributes of all the pottery found at the Palmetto Grove site follows. This list is not exhaustive, but represents those features which seem most favorable for cultural analysis,

1. Ceramic Chunks, With Temper. These are chunks of clay, similar to the clay used in making the1 Palmetto

Ware pottery, including the shell temper. These lumps were first taken to be simple blobs of prepared clay.

It may be they were actually cylindrical in shape, and were once rolled lumps of clay being saved as raw material for making pottery. Columbus may have seen just such a cylinder of clay on 15 October 1492 (Morison 1942; 240).

After having left San Salvador he reports picking up a native in a single-man canoe who carried with him Ha lump of bright red earth powdered and then kneaded."

2. Ceramic Chunks, Without Temper. These may be lumps of raw clay to which temper has not yet been added. 51

There is no temper and absolutely no continuity in shape from one chunk to the next,

3. Red Coloring Matter in Clay, This color is apparently derived from the red Pineapple Loam soil to be found on the island. All of the pottery referred to as

Palmetto Ware is this color.

4. Buff Color Paste. Vessels apparently fired in an oxidizing atmosphere, producing a light brown, "buff" colored paste,

5, Shell Temper. Crushed shell, not just broken fragments. Always found in Palmetto Ware pottery.

6, Grit Temper. Small grains, frequently quartz.

7. Bowls. A general term applied to all rims that do not represent constricted necks, or which are not from griddles.

8, Jars, A general term, applied to all vessels with constricted necks, «

9, Griddles, All griddles appear to be rimless, that is, pancake type (Hoffman 1963s 64).

10. Red Slip. A slip or paint apparently made from Palmetto Ware. In the five sherds with this trait the slip was applied rather heavily. Like Palmetto Ware, it also contains shell particles.

11. Flat Rim Lip (Fig. 8).

12, Rounded Rim Lip (Fig. 8). 52

figure 8. Ceramic Attributes on Pottery from Palmetto Grove.

A-rim points, B-lunar, wedge-shaped lug, C-incised rim, ID-flat lip, with roll, C-Incised sigmoid design, F-incised lug, G-strip lugs, H-zoomorphic head design. 53

13. Flat Rim Lip, Exterior Roll (Fig„ 8) „ The

lip is rolled about 1 or 3 mm, away from the top of the vessel, making the rim top wider than the thickness of the vessel wall,

14. Rim Points or Horns. Where a normally curving rim is pinched to a point. These may project out

5-10 mm. There may be a single point or two close together (Fig, 8),

15. Thick Walls, Eight mm. and above (usually associated with Palmetto Ware).

16. Thin Walls, Less than 8 mm. thick (usually associated with pottery wares other than Palmetto Ware).

17. Basketry-Impressed, Checkerwork.

18. Basketry-Impressed, Twilledwork.

f 19. Basketry-Impressed, Wickerwork.

20. Basketry-Impressed, Twined,

21. Incised Rims (Fig, 8), Incising across the top of the rim. In one case this occurs directly above a wedge-shaped lug.

22. Lug. All projections outward from the side of the rim or side of the vessel. Some may be incised.

23. Crest Lugs, A term used by Rouse (1952$ 345,

PI. 4 I) to refer to a strip of paste draped across the rim of a vessel (Fig. 8).

24. Lug, Modeled-Incised„ Possibly zoomorphic. 54

25. Wedge-Shaped Lugs, Appear wedge-shape in cross-section, Applied to the side of the vessel, below the rim, with the long axis parallel to the rim of the vessel„ One is lunar-design (Fig„ 8)„

26, Incised Sigmoid Applique''. A single applied strip, shaped like a gently curving sigma, lying on its back. Every few millimeters it is incised at right angles to the strip (Fig. 8).

This list summarizes the attributes observed on the pottery at the Palmetto Grove site. They are spread over the four kinds of pottery mentioned above. The list is helpful in many ways, but particularly in attempting to plot the distribution through time and space of minimal elements of cultural patterning. An excellent example is the incised sigmoid design (No. 26). This particular design is found in a number of pottery stylings in the

Greater Antilles after A. D. 800. By isolating this phenomenon from its cultural matrix on San Salvador we are able to pursue it, perhaps establishing relationships, or revealing hints of relationships with other peoples of the Caribbean.

This situation points up the utility of Rouse's mode analysis as opposed to type.classification. The incised sigmoid design appears on the buff, thin, sandy- paste pottery. If I were to provide this pottery with a 55 type name based on the association of a number of non-= comparable attributes, and then attempt to trace its development as a type, the significance of the design element would be lost. It must be emphasized, however, that the occurrence of the same attribute in two different sites does not mean that there is necessarily a relationship between the two. In my opinion these are only hints; they must be pursued to a more general level of comparison before real relationships may be estab» lished. The type may not be the proper level, either, but it would seem that the simultaneous clustering of the same modes into ceramic styling at two different locations would not be likely to happen accidentally. Relationships must be pursued at the level of assemblages. An assemblage is the entire collection of attributes of a given component of a site, such as the assemblage listed above. If we were to find an identical listing on another island it would seem logical that the two islands were in contact with each other. When dealing with more complicated ceramic assemblages, this contact should be all the more certain.

It remains now to show what modes are associated with Palmetto Ware and other pottery wares found at the

Palmetto Grove site. 56

Palmetto Ware

Those attributes which characterize the * Palmetto Ware pottery are listed in terms of three basic non~comparable modess paste, form, and surface„

1„ Paste

Colors always red

Tempers always shell, may also include beach sand, may also include clay lumps

2, Form

Shape

Bowls Jars Griddles

Rim

Flat lip Round lip

Thicknesss always thick, 8 mm, or more

3, Surface

Colors always red

Finish: plain and smooth

Alteration

Basketry impression

Checkerwork Twilledwork Twinedwork . Wickerwork

Crest lugs

Incising across rim

Rim points 57

In summary Palmetto Ware is always red and always

shell-tempered, but may be in the form of jars, bowls» or

griddles, and may have a rounded or a flat rim lip.

Occasionally basketry impression, lugs, incised rims and rim points may occur.

"Trade" Wares

The three remaining kinds of pottery found in the

Palmetto Grove site may or may not be "trade" wares.

Temper in the paste,of some of these sherds is not naturally found on San Salvador. However, I do not know if the tempering materials themselves, the prepared paste, or the completed vessels were introduced„ Two of these three varieties of pottery are tempered with a fine sand composed of quartz and feldspar, and one also contains hornblende. They share many characteristics; modes for both are listed below.

1. Paste

Color

Buff to brown Reddish gray <.

Temper

Fine sands quartz and feldspar, may also include feldspar

2. Form

Shape 58

Jars Griddles $ only one griddle was found that is not Palmetto Ware; it is pancake type with reddish-gray paste

Rim

Round lip Flat lip Flat lip with exterior roll

Thicknesss all thin, 7 mm„ or less (except for the one griddle sherd)

3 o Surface

Colors always same as paste

Finishs plain

Alteration

Lugs

Crescent-shaped Zoomorphic head Semi-circular

Incising

Incised sigmoid applique.

The third possible trade ware variety consists of

five sherds. They are sufficiently distinct that they may represent a hint of pottery yet to be fully appreciated from the Palmetto Grove site. These thin sherds are made

of a soft black paste that is tempered with shell. The vessel surfaces have a thick red slip that also contains minute bits of shell temper. 59

Comparison tilth Other

tiest Indian Ceramics

The Palmetto Grove is the only site yet reported

on in the Caribbean with a ceramic assemblage such as

described above. This distinction is partly a matter of

a lack of investigation, however, as it was the first site

in the Bahama Islands to receive serious archaeological attention.

The bulk of the pottery in the New World Museum, apparently all from sites on San Salvador, is tehe same

thick red Palmetto Ware, except that some of it is

decorated. The designs on this pottery tell more of a

story than the plain ware from the digging of the Palmetto

Grove site, although in the letter's meager display of attributes we still see cultural affiliations to the south.

The pottery in the Museum combines both Meillac and

Carrier traits, two styles of pottery found on Haiti by

Rouse (1939, 1940), These are perhaps the two closest wares to the Palmetto Ware.

On the basis of a survey of the ceramics.

Cranberry concluded that there were three stages of

cultural development in the Bahamas, In the following discussion I shall borrow freely from his 1955 and 1956

papers.

Of the 61 major sites he reports, only 16 represented open villages. The remaining are all cave 60 sites, Productive sites are concentrated in the Caicos in the southern Bahamas, rather than the north.

Fifteen of the vilJLage sites are found there. Ceramic and non-ceramic specimens recovered from these sites show marked similarities (Cranberry 1956: 129) to artifacts from north Haiti as defined by Rouse (1941:

54-91, 113-40, Pis. 7-24, 27-34). This is particularly true of the two dominant ceramic styles: Weiliac and

Carrier„

The Meillac series in both Haiti and Jamaica apparently developed out of a local series called

Ostiones (Rouse 1963 s 509), the type site for which is located on the west coast of Puerto Rico. The Meillac style (Rouse 1939: 42-3) is a rather hard, thin, red ware, averaging between 3 and 7 mm. in thickness. Jar shapes are rare. Characteristic incised designs include cross- hatching and a series of oblique parallel lines, consisting of alternatively inclined units. The Weillac series survived only in Jamaica, central Cuba, and apparently the northern Bahamas.

The typical, reconstructed Weillac specimen

(Cranberry 1956: 129-30) is a vessel, probably a bowl of moderate size with a large aperture and may have been round or boat-shaped. The bowl sometimes bears a red clay

) slip and the Walls are thick and only moderately polished. 61 The surface of the vessel is hard, and there may be a narrow shoulder, turning not far below the lip. The lip itself is either round with a straight rim or beveled with a slightly flaring rim. The surface is usually not decorated, though in some instances it may have an incised design on the shoulder just below the lip, usually a crosshatch design which extends around the vessel, Lugs occasionally occur in the form of an apimal9 s face.

Carrier style of pottery succeeded the

Meillac. Its makers were the Indians found by Columbus when he discovered the New World. This pottery (Rouse

1939 $ 43) is thicker» with softer surfaces, and a grayish-brown clay color. Jars are common. The edges of the rims are usually flat and lugs are large and modelled. Incised lines are broad and shallow and spaced more widely„ Lines commonly terminate in dots,

Characteristic incised designs include a series of lines curving in and out among each other, and panels, oval in shape, inside of which is a line terminating in dots.

The typical reconstructed Carrier Specimen from the Bahamas (Cranberry 1956s 130), is a thick-walled bowl, the exact body shape and bottom form of. which are uncertain. It is, however, probably round or boat-shaped„

The surface is soft in comparison with Weiliac pottery.

Lips are round, and shoulders rarely occur. Decoration, 62

though rare, consists of curvilinear rather than straight

lines as on Weillac specimens„ Zoomorphic head lugs . representing the features of a bat are common, usually on

both ends of the vessel. Carrier specimens never have a slip. On the whole these sherds are of finer quality workmanship than are Meiliac sherds.

In addition to these predominant ceramic styles,

Cranberry (1956s 130) discovered several others, including one similar to Palmetto Wares

One is defined from numerous body sherds from the central and southern islands. They are probably from round bowls and are crudely made, soft, and friable„ They are characterized by fabric impressions from a twined material. The style is well repre­ sented and, in the Caribbean, is unique to the Bahamas.

Cranberry (1956, Fig. 1) places San Salvador at the northern edge of the "central islands" of the Bahamas.

In attempting to reconstruct the historical position of Palmetto Ware one quickly is made aware of the futility of working with the modes of this particular pottery, let alone trying to do something with it as a type. However, despite the lack of decoration on the excavated material, it is possible to see that at least the Palmetto Ware possesses a few attributes that seem to be a part of Me iliac and Carrier styles. Referring to the listing of characteristics of Palmetto Ware pottery we also can see a closer affinity to ffleillac than to Carrier. The Following list identifies the modes for these two styles (Rouse 1939s 55=6) that can be observed on Palmetto Ware materials

Mode Meillac Carrier

Red Color Common Rare

Jar Rare Common

Flat Top Rim Rare Common

Wedge-Shaped Lug Present Absent

In addition, modes on the possible trade wares from the Palmetto Grove site that occur on either Meillac or Carrier are:

Mode Meillac Carrier

Wedge-Shaped Lug Present , Absent

Zoomorphic Head Lug Rare Common

Sigmoid Design Common Rare .

Ridge on Outside Rim Common Common

These lists might make it possible for one to conclude that Palmetto Ware was nothing more than the San

Salvador version of late Meillac, However, knowledge of

Palmetto Ware from other sites on the island points up more of the Carrier influence. In particular, a site near

Palmetto Grove called the Ward Site where Goggin excavated produced a high percentage of jars. 64

Cranberry (1956) has pointed out a definite scheme

of distribution of ceramic styles in the Bahamas. Meillac

sherds are found from Great Abaco south through Grand

Turk. Carrier specimens are found only as far north as

San Salvador, Rum Cay and Long Island. Palmetto Ware,

which has affinities for both Carrier and ffleillac styles

occurs only as far north as Cat Island. On the basis of this Cranberry (1955, 1956) divided the Bahamas into three zones, the central one being an area of overlap, or transition.

Cranberry assigned the islands from Grand Bahama and Great Abaco south, probably including Great Exuma and

Cat Island, to the northern sub»area. The Meillac style was apparently in continuous occupation there until the extinction of the Bahaman Indians in the early and middle 1500's.

The islands from Cat Island, San Salvador, Rum

Cay and Long Island south to Great Inagua and the Turks and Caicos Cranberry called the central or transitional

sub=area. The Meil-lac style was there for some time, but

Carrier influence was beginning to make itself felt by the time of European intervention.

Great Inagua and the Turks and Caicos constitute a southern sub-area. The Meillac style was replaced almost in toto by the Carrier at some time not long before

Columbus appeared on the scene. 65

The Palmetto Ware may turn out to be a kind of marker for the transitional period, from Meillac to

Carrier. Future work in the central sub«=area should bear this out„ Palmetto Ware sherds have been found in some quantity on Cat Island during Florida Atlantic University surface surveys in the winter of 1965-66. Three sites, including one large village site, were located on that island. The bulk of the pottery found on the surface is

Palmetto Ware.

During archaeological reconnaissance in the

Bahamas in 1936 Krieger (1937s 98) discovered a pottery similar to Palmetto Ware on Long Island.

The pottery recovered by the ex­ pedition from a Lucayan site at the Hamilton caves in the Deadman's Key district of Long Island appears to be typical of finds made elsewhere, Hand-molded figurine heads which occur as handles joined to the earthenware bowls of the island Arawak were not seen there but were noted on the thin-walled red ware pottery recovered from the Salt Pond Hill cave, Inagua. The prevalent decorative design on the thick-walled, shell-tempered Hamilton Cave pottery consists of incised recurved volutes, nucleated circles, and incised lines broken by isolated punctations.

Thick walls and shell tempering are two of the diagnostic attributes of Palmetto Ware. Unfortunately

Krieger does not illustrate this pottery, nor does

Cranberry. It may also be possible that the thin red ware forms from Inagua are the same as the thin red ware from San Salvador. The incising on the thick pottery 66 sounds more like Carrier than Meillac (Rouse 1939s 55, Pis,

IV 6, V 4, 5), If this thick pottery is the same as

Palmetto Ware, then we could conclude that the latter is indeed a marker for the transition from Meillac to

Carrier,

There is yet to be considered one disturbing point,

One would assume that if the center of the sphere of cultural influence was really the island of Hispaniola, then the Bahamas might naturally be referred to as a sub-culture of the larger island„ And, excepting the thick, shell-tempered red ware, this is what seems to be the case, But, when an effort is made to track down the modes on Palmetto Ware, the focus is not on Hispaniola, but rather the VirgintiIslands, far to the east„ It does not take very long for one to become convinced of the similarity in attributes between Palmetto Ware and Virgin

Islands pottery. Simply thumbing through Bullen1s paper

(1963) on the ceramic periods of St„ Thomas and St, John in the Virgin Islands reveals unusual decorative features of the Magens Complex such as rim points, horn-like projections either single or paired, sigmoid designs, incised strip appliques, vertical strip applique, incising

, I ' at right angles across the rim, crescent-shaped, wedge- profile lugs, and even the presence of a reddish paste,

All these attributes may be found on Palmetto Ware pottery. 67

There is apparently a difference between Palmetto Ware

paste and that found on the Virgin Islands* However, this difference probably means no more than that a similar

style of pottery was locally made in both island groups*

Similar modes are found on the Santa Elena style pottery in eastern Puerto Rico (Rouse 1952s 344, PI* 4),

With Santa Elena, although the material is different, apparently even the construction is similar (Rouse

1952s 352).

Santa Elena pottery is relatively crude, the majority of the sherds being coarse and fairly soft. * *

The sherds are the thickest in Porto Rico, They average 8 mm. in thickness and some are twice that amount* Nevertheless they break easily. The fractures are coarsely granular and, in some cases, dis®- integrate when rubbed with a finger. For the most part, they are reddish brown in color.

In addition, Santa Elena sherds have crest lugs, vertical applique', vestigial handles with vertical parallel lines incised on the handle. Sigmoid designs are rare, Rouse, whose work was much earlier than Bullen1s, came to the conclusion that because Santa Elena is

"restricted geographically to Vieques Island and the eastern half of the main island" it probably developed first in the Virgin Islands and spread from there to the adjacent parts of Puerto Rico. 68

The implication is clear that the Bahamas

Palmetto Ware is related to certain pottery styles in

the eastern Caribbean, but not necessarily to the place

of closest contact, Hispaniola.

It appears, thus, that at least as far as San

Salvador is concerned (and I suspect that further

research will expand the area) we are faced with three,

not two, major ceramic styles: Meillac, Carrier, and

Palmetto Ware. If we confined our interpretations to

the Palmetto Grove site, we would have to drop the first

two. Indeed, most of the pottery that I have observed from ovef the island of San Salvador, or from Cat

Island, is Palmetto Ware. It is only because of a few

partly reconstructed vessels (of whose context we are

uncertain) in the New World Museum, and some support from the literature, that it can be inferred that Meillac and

Carrier are there at all.

One of the more important single sherds to be recovered is a thin, buff, sandy-paste ware with an

incised sigmoid applique' design. This particular design occurs in the Ostionoid Style (coeval with Santa Elena) of western Puerto Rico (Rouse 1952, 1963), in the Magens

Complex of the Virgin Islands (Sullen 1963, PI. 2 C), and on Jamaica (according to Ronald Vanderwal). The time is all post-A. D. 700 and the ceramic horizon is generally 69 presumed to be the predecessor for the Meillacoid Series of the Greater Antilles,

Comparison With Mainland

United States Ceramics

Although the material recovered from the Palmetto

Grove site does not resemble anything reported from the mainland of the United States„ the possibility remains that there may have been contact between the people of the Palmetto Grove site and Florida.

The Indians Columbus encountered indicated they had been attacked by people to the northwest, but we have no knowledge of whether this was another island to the northwest, or the mainland. Both Arawaks and Caribs are reported ethnographically (Roth 1924) to have been constantly carrying on inter-tribal warfare; perhaps such was the case from island to island in the Bahamas.

The Bahamas are the closest islands of the West

Indies to Florida and it would seem that if contact had taken place the evidence would be found there. With this in mind Cranberry (1957) searched several islands close to the mainland for contact sites, but found none„ There have been many talks and conferences and several lengthy papers covering the subject. The general consensus seems to be that there is little if any relationship between

Arawakan groups and the Florida Indians. It is perhaps 70 unfortunate that nothing Was found in the Palmetto Grove site to help resolve the problem* On the other hand, the lack of evidence for contact provides further I support for stating that there was no relation at all. CHAPTER 4

OTHER ARTIFACTS

Most of the artifacts from the Palmetto Grove site are divided into classes or groups. Each group consists of a single class of artifact. The groups include pottery, shell scrapers, coral abrasive implements, shell ornamentsp shell gravers, shell gouges and food shells.

Further, some shell and coral objects are also classified into types. It should be but a short step to go from there to modes, if it is considered necessary in future research.

Stone

Specimens vary from objects introduced by man, but which apparently served no particular function them­ selves , to artifacts of worked stone.

Man-Introduced Stone. This category covers stone that has not been shaped or worked or even apparently used. These objects are singled out for reference because they are not considered to be native to San Salvador.

Fragments includet strontianite, calcite, chert, gneiss, limonite, feldspar1 and quartz. 71 72

Water Trough- Along the south edge of Section

0-7 was found a huge stone vessel of natural limestone.

It is similar to naturally, shaped bowl-like concavities formed in the rocks at the rocky point adjacent to the site„ The vessel (Figs. 5, 9) measures about 60 cm. across the top, and is 44 cm. deep in the deepest part.

The bottom is rather flat and rested at 28 cm. below the surface of the ground.

This object is rough, not 11 finished." One man can lift it with some difficulty, but it takes two men to carry it more than a few meters.

Sam Ferguson, resident of the island working on the digging crew, said the local people now use these formations for water for chickens and call them "chicken troughs.11 Another crew member said he had always called them "hog troughs„"

Just to the east of this bowl there was a dark lens of charred wood which turned out to be the burned root section of a palm similar to the Sabal palms in the area„ It would seem that if the two are temporally related that the bowl had been placed in the shade of the palmetto.

Limestone Slabs With Shallow Grinding Depressions

(Fig. 10 H ’ I). Two limestone slabs, one approximately

3 cm. thick and the other 5.5-5.8 cm, thick, were found 37cm------1 "1 E o J

Figure 9. Limestone Water Trough. 74 with slight depressions in one surface. These depressions might have passed as natural if microscopic analysis had not shown that the surfaces had actually been ground down.

The limestone is composed of cemented calcareous organ­ isms such as oolites, foraminifera, and small shell fragments.

The function of one slab is apparent from the chunks of red material found ground into -the matrix.

Touching this red material with a dilute solution of hydrochloric acid produced no effervescence, although the matrix surrounding it dissolved. This slab was probably used for grinding paint. No direction of grinding movement is evident.

The second slab contained no foreign matter, but in general presents a darker, almost smoke-blackened appearance. Uni-directional grinding facets are easily detected when intense light is played at a low angle across the depression. These grinding facets are numerous, but shallow and with smoothed edges.

Both grinding slabs were found in the 20-30 cm. level, but 22 m, apart. It is interesting that the one without evidence of paint was found in association with another ground stone with a very deep depression, shaped very much like a metate (Fig. 10 F ), but of stone much too soft to serve as a corn grinding implement. Near these two objects was a heavy concentration of pottery.

Figure ith Stone Tools„

A-C, Coral rasps (B is reverse side) D, Coral corn shucker; C , Coral file F, Grinding stone; G, Coral hone; H-I, Limestone slabs with shallow grinding depressions. 75

Stone Tools 76

All this was along the eastern edge of the excavation, while the other shallow grinding slab came from along the western edge.

Along with the red paint stone were found many small chunks of limestone, about 3 cm, across, some of which had fire marks. They may have been introduced from a nearby charcoal or lime-making pit discussed elsewhere in this report.

Grinding Stones. Only three fragments of this artifact are available. They come from three different grinding stones. In profile (Fig. 10 F) they are flat on the bottom with the upper surface having a smooth trough­ like depression in the center. The matrix is similar to that for the above-mentioned Limestone Slab With Shallow

Grinding Depressions, except that the basic material is far more angular and contains much more crushed shell.

The curious thing about these artifacts is they are apparently too soft to serve as metates in the sense that they were used to grind corn. 1 suspect that they wdre used for grinding something softer, such as manioc root, or berries. Although the shape is not unlike that that of a griddle, it is doubtful they served in this man­ ner as there is no indication of their.having been in a fire. The upper and lower surfaces differ considerably.

The lower, or bottom, surface is ground as though the 77 object had been placed on a hard surface when used. The upper or apparently functional surface presents an even grinding plane, but is really rough in texture. This rough surface would be ideal for grinding or peeling root crops.

All fragments have been found within 6 m. of each other, two at the 20-30 cm. level and one at the 10-20 cm. level, immediately above another.

Hammerstone-Planer. One hammerstone-planer was found. It measures 7.2 cm. wide by 9 cm, long and is 4.3 cm. high. It is unifacial in that one surface is convex, the other flat. It was apparently pecked and ground into shape. The material is strontianite, and may not be native to San Salvador.

Evidence of use occurs on both the convex and flat surfaces. The convex surface is battered and shows an occasional scratch? the flat surface has been used as a planer-seraper. Striations are easily detected with the aid of a microscope. Although shaped very much like an

Archaic unifacial scraper, its use was apparently as a hammer and a planer.

In association with this hammerstone-planer were several shell tools, including one Strombus hammer and a few Strombus gouges. 78

Beads. Three stones that can be called "beads" and two that are apparently bead "blanks" were Found; one of the former is questionable.

One bead is barrel-shaped (Fig. 11 Y), ground from an igneous rock. It is 10 mm, high by approximately

14 mm. in diameter. The perforation is 4 mm. wide at one end and 2 mm, at the other. It was drilled from the former end and when very close to going all the way through was drilled From the other. The drill hole is neat and round. The second bead is a fragment, split down the middle, the height beingat least 13 mm., and the diameter of the drill hole, 4 mm. Neither bead is calcareous.

The bead "blanks" are again barrel-shaped, ground igneous rock, probably feldspar (Fig. 11 U). The beginning of a perforation may be seen in one of the flat surfaces of each specimen.

The questionable bead (Fig. 11 X) is a fragment of limestone with a neat hole drilled through the center. It is 14 mm. high, 13 mm. across its widest part, and with a

2 mm. diameter perforation.

Problematical Stone Objects. Five barrel-shaped stones and one in the process of being shaped (Fig, 11 S ,

T, V, Ul) of igneous rock (probably feldspar) were found in the southern half of the site. They may be bead

Figure 11» Shell» Bone and Stone Ornaments,

A-I, L, Shell beads § J , K , 0-K, Shell tinklers; N , Bone ornament; 5, I, U , W, Barrel=>shaped stones; U» Bead "blank;" X , Y, Stone beads. F G E

R N m 9 0 ST U V Shell, Bone and Stone Ornaments.

^ 3 m vO 80

"blanks," but as they outnumber the finished product,

they just as well may not„ They could have been part of

a game, or for ceremonial use. Their measurements

follows

Excavation Heiqht Diameter Level

25 mm.* 16 mm. 0-10 cm.

15 mm. 16 mm. 10=20 cm.

12 mm. 15 mm. 10-20 cm.

14 mm. 16 mm. 20-30 cm. 15 mm. 16 mm. 20—30 cm. 13 mm. 14 mm. 20-30 cm.

* There is the beginning of a cut at 16 mm. from one end.

A second problematical stone object is a fragment

of a highly polished» gray-blue, fine grained rock.

Probably a gabbro, this fragment was apparently part of a stone celt, such as typically found in the Greater

Antilles.

A third problematical stone artifact is a cone-

shaped object carved from strontianite. It is blunt on

both ends and has a slight groove 5-8 mm. from each end.

Microscopic examination revealed numerous fragments of

crushed red pigment ground into both of the ends.

Apparently it served to crush red paint from hematitic or

limonitic rock. It was recovered from the 20-30 cm. level,

a few meters away from the palette-like grinding slabs. 81 Another object„ a flat D-shaped pebble, apparently

calcite, has no grinding facets and no battering marks.

Its use must remain unknown at this time and it is only

mentioned because the material is foreign to San Salvador,

A loaf-shaped calcite pebble was found. It is i 2.7 cm. long, 2.2 cm. wide, and 1.3 cm. high. It is

blackened, apparently by fire. It is possibly a pot

scraper, but no striations or scrape marks are visible, even under the microscope.

One ground stone pebble is apparently a feldspar,

but it also contains a crystal of olivine. More or less

loaf-shaped, it is not as turtle-backed as the pebble last

mentioned. It has a few tiny fragments of red pigment ground into scratches and crevices that are noticeable

under the microscope. However, this object is so small

that its function is hard to interpret. It measures 7 mm,

by 16 mm. by 19 mm.

Two other stone objects must be mentioned, although it is questionable as to whether or not they are artifacts. These are calcite fragments, both about the

same size and both bearing marks possibly made when used

as hammerstones. One resembles one-half of a loaf-shaped

turtle-back scraper or hammerstone. It measures 2.2 cm.

high by 4.1 cm. wide and is presently about 2.6 cm. long.

Its present shape suggests that it was probably slightly

over 5 cm. long when whole. The other object is pie-shaped, 82 the worked portion being around the curved edge. One

leg of the pie is 4.2 cm., the arc is about 5.2 cm. around, and the height is about 9 mm.

Bone

Only one bone artifact was found. It has been perforated (Fig. 11 N) in three places, perhaps for stringing onto a necklace. The holes seem too small for insertion of feathers. There are two holes in one end a the lower end in the photograph.

Shell

Shells are by far the most numerous class of artifacts recovered from the Palmetto Grove site. This includes shells of many varieties and uses, that is gouges, gravers, hammers, choppers, scrapers, ornaments, weights, dippers, a trumpet and, of course, residue from meals.

Without doubt many of the functional categories listed here will evoke negative comment from colleagues engaged in anthropological research in the Caribbean.

Some have already expressed conservatism in personal communication. The truth of the matter is that we simply do not know enough about shells as part of the pre­ historic technology. We do not really know what shells can do or be made to do when the people do not have stone with which to work. Supposedly all the non-limestone raw 83 material on San Salvador has been imported. The native limestone is relatively soft. The first logical con­ clusion then would be that the inhabitants would have used abundantly available shell to stock their inventory of tools. Where we would normally expect chert scrapers and choppers, we should look for shell' scrapers and choppers, not simply ignore the possibility of their existence. It does not follow that all cultures have to have the same tool kit, but it should be expected that certain basic tools may be found, and if they are needed they will be made with the best available materials.

Studies are needed on the strength, working, and use of shell, and tests to demonstrate what the products should look like if they have been used. To my knowledge, however, nothing of this kind has been done on shell in the Caribbean, One of the reasons for boldly stating that certain items served as tools is to stimulate discussion and perhaps a better understanding of the nature of shell in the cultural assemblage. Shell items must be studied from a functional point of view. They must be analyzed in terms of fractures, breaks, wear marks and cuts, and they must be approached with some idea of the types of tools the particular culture would demand, such as fishing implements or farming tools. An example is the shell gouge formed from the conch (Strombus qiqas) shell, roughly triangular in shape, pointed at one end, and broadening 84 out along the natural- lines of the conch. The broad, curved, upper end is in most cases knicked an,d fractured, as though it had been used for gouging such things as wooden items, or perhaps from use as a hoe or digging tool. The function was much the same as the present-day chisel. These gouges clearly seem to have been used for some purpose. The knicks and fractures resemble "use” marks common on flint or chert tools.

Another implement called a gouije is already described in several reports on the Caribbean, It possess­ es certain recurring criteria, one of which is that it must have a ground and bevelled working edge (Rouse

1960, Fig. 10 A). Because the shell implements described below do not have a ground edge, it has been suggested that they are not gouges. Two thoughts occur here:

(1) a difference in function could result in the used end being broken instead of ground, and (2) the San Salvador inhabitants probably could not have ground the edges if they wanted to because there is not a native rock on the island that is harder than the shell involved.

Shell Scrapers. These are exactly what their title implies, shells that have been used for scraping.

One of the present-day San Salvador uses of the Telllna scrapers is for scaling fish. There are several different kinds of shell scrapers, including Codakia. Lucina.

Tellina. and two specimens from the family Tellinidae, S Figure 12. Shell Scrapers. A-L, Q „ Tellina scrapers; N-P, Lucina scrapers; M , R , S , U , Codakia scrapers; I, V , Codakia scraper-graver. 85

Figure 12. Shell Scrapers. probably the genus Arcopagla. Including the scraper- graver , there were 543 whole specimens found; 310 Codakia scraper fragments were also found. 1. Codakia scrapers (Fig. 12 M, R , 5, U).

Codakia sp. valve, probably from C_. orbicularis.

Vary from approximately 4.3 to 8.1 mm. in length.

Crenulations on outer edge are completely

worn away. Outer edge normally complete.

Seventy-one whole specimens were found.

2. Codakia scraper-graver (Fig.

12 T , V ). Codakia sp. valve, probably jC.

orbicularis. Differs from Codakia scraper

in that the outer edge is not complete, but

is abruptly broken off at mid-point, pro­

ducing a sharp corner, similar to that on

a graver. This tool type may simply be a

broken Codakia scraper and may eventually be

lumped into that category. The 23 Codakia

seraper-gravers found came from the top

four levels of excavation, most from the

20-30 cm. level; none were found below 30-40 cm.

3. Lucina scraper (Fig 12 N-P). A

valve from the L_„ oensvlvanlca„ a small

but thick shell. Outer edge is completely

worn, but not always all the way around. Table 6. Vertical Distribution of Shell Scrapers.

Only 11 squares were excavated to 50 cm., 5 to 60 cm. See Figure 12.

Excavation Levels in Cm. Shell Scrapers T ota 0 = 10 10=20 20=30 30=40 40=50 50=60

Codakia so. 5 16 40 9 1 71

Codakia fraoment 36 107 117 36 11 3 310

Lucina oensylvanica 3 5 10 1 19

Tellina so. 4 5 4 13

Tellina so. 1 8 22 6 1 38 Left Handed

Tellina sp. 7 6 27 10 3 53 Right Handed

Arcooagia so. 1 1

Unclassified 2 3 6 4 15

T otal 54 150 227 70 14 5 520 88

These shells measure 22-28 mm. long and as

much as 5 mm, thick. Nineteen were found,

4. Tellina scraper, left handed

(Fig, 12 D, G, H, L , Q), A valve from

Tellina sp. shell. In all cases so far

observed this type of scraper has two ob­

servable characteristics: (l) all are left

valves, (2) breakage is always to the right

or anterior portion of the shell when

viewed on the inside of the shell. Size

varies from 4.8 to 8.9 mm. Thirty-eight

were found.

5. Tellina scraper, right handed

(A-C, E, F , I-K). A valve from Tellina sp.

shell. In all cases there is one observable

characteristics breakage is to the left

when viewed on the inside of the shell.

Shells vary in length from 3.7 to 6.6cm.

Fifty-three of these were found.

Shell Graving Tools. Many shells have been cut and ground into a specific shape (Fig. 13) that is approximately repeated from fragment to fragment. The : ■ V most consistent aspect of this shape is the point at one end. This point has been worn, chewed up from use, and occasionally broken off. In some cases no use marks appear but the shell has been cut and ground into a shape % Figure 13. Shell Gravers and Drills. A, B, E „ G p H, J, P, S-X, BB, are Type A gravers; D p F , R , Y , Type B gravers § C , I, K-0, Q , Type C gravers; BB=EE, Drill, Type A; FF, GG, Type B Drill; AAf Type C Drill. 89

Figure 13. Shell Gravers and Drills. 90 Table 7„ Vertical Distribution of Graving Tools.

Only 11 squares mere excavated to 50 cm, 5 to 60 cm. See Figure 13.

Excavation Levels in Cm, Graving Tools T otal 0-10 10-20 20-30 30-40 40-50 50-60

Type A

Strombus 13 28 18 70

Triton 1 1 2

Codakia 1 1

Unclassified 7 16 3 26

Sub-Total 13 36 35 13 99

Type B

Strombus 2 3 1 12

Triton 1 1

Sub-Total 3 3 1 13

Type C

Strombus 3 1 10

Triton 1 1

Unclassifled 1 1 2

Sub-Total 3 3 5 13

T otal 19 39 43 21 125 91

Table 8, Vertical Distribution of Shell Drills Cut Frodi Strombus qioas.

No drills were found below 40 cm. See Figure 13.

Excavation Levels in Cm. Shell Drills — ------— ---- — --- Total 0-10 10-20 20-30 30-40

Type A 1 2 1 4

Type B 2 2 3 7

Type C 1 2 2 5

T otal 1 5 4 6 16 92 similar to those on which use marks do appear. There may * have been many more of these tools than reported here that went unnoticed because of their fragmentary condition.

Many (Type A for example) are formed in such a way as to produce a "backed point," that is, as if the point were to be pulled or pushed along a surface, rather than drilled through it. The tip would not readily be broken off. At the same time, the point is small enough to have been used as an awl.

It is not clear what the implements, called

"gravers" here, were specifically used for. Perhaps they were used for incising wood, shell, or pottery, for punch­ ing holes in cloth, pottery, or as a fid in making baskets.

A total of 125 was found in the Palmetto Grove site.

Gravers are divided into three alternative types:

Graver, Type A (Fig. 13 A, B , E, G, H, J,

P , S-X, BB). These gravers are flat. In profile

they may be shaped as a right-triangle, or

curved.

Graver, Type B (Fig. 13 D , F , R, Y).

This graver is formed from the horn of a

young conch. In profile, looking at the

"front," it is triangular, looking at the

side, it appears as an upside-down L. The

little arm projecting out was apparently

used to gain leverage. 93

Graver, Type C (Fig. 13 C , I, K=0, Q, A).

This graver is flat and is often shaped like a

triangle, having been tapered to a point along

both edges.

Shell Drills. Certain shell fragments have been separated out because their shapes suggest they were used as drills. In many cases this means they are simply long and come to a point. But it takes more than this to make a drill and they must be examined carefully to see if the point has been worn or scratched» This usually must be done with a microscope and directed lighting„ There are three types of drills,•Types A , B , and C . All but two were recovered from the 20-30 cm. level. One of the Type

A drills was found in the 30-40 cm. level. The Type C drill was found in the 10-20 cm. level, slightly above one of the Type B drills.

Drill, Type A. This drill (Fig. 13

BB, CC, DD, EE) is fashioned from the whorl

of a conch shell. It is concave-convex„ There

are two specimens, the tips of which are ground

off (Fig. 13 CC for example)„ All were apparently

cut into shape in the same manner because they

exhibit similar cutting and grinding marks.

Drill, Type B. (Fig. 13 FF, GG). These

are simply long, thin cut fragments of shell,

probably conch. The points have been used. 94 One tip has been broken off„ All have been

cut with the cut edges ground smooth. There

are seven of this type drill.

Drill, Type C (Fig. 13 AA). This

drill is shaped very much like its chert

counterparts. It has been cut from a conch.

It not only exhibits use mgrks at the point,

but has apparently been handled many times

by wet or sweaty hands as the top edges

have been worn smooth.

Shell Weights. For a long time it was difficult to discern shell weights. Weights were looked for be­ cause it is believed that one of the means of obtaining fish was by netting them, or by using submerged basketry traps. Some fish are algae eaters, the parrotfishes for example. As over half of the bones recovered represent algae-eating fish, it seemed logical to expect to find some means of catching them. If nets or traps were used then there should be weights? if spears, then spear points unless the latter were tipped only by hardening the wood.

The first clue to a possible net or trap weight came after examining the outer whorl of an apparently useless conch

(Fig, 14 A) in which there is a round perforation. The hole is not in the proper place for hafting, nor for cut­ ting the muscle of the animal in order to extract it from the shell (such as the hole above the outer whorl in UlNO Figure 14. Strombus Shell Weights. 96 Fig. 14 A). A second shell with only the apex and part of the cdlumnella remaining has a similar perforation (Fig.

14 C). This perforation is slightly worn along the outside edge. The hole was first made for removing the meat, but had been used later by passing something through it, perhaps a length of line, Both shells, which were found next to each other, were then examined and it could be seen where on the shell a line may have been passed. On the whole conch specimen, there is a slight chipping along the lip of the outer whorl„ In fact the whorl had been broken partly away, just as one might expect from continuous use as a weight. The same slight retouching was noticed on the second shell along the outer edge of the apex fragment.

I infer that these shells were used as weights either for fish nets or fish traps. We have no direct evidence for determining which method was used for catching fish. However, algae-eating fish live chiefly on reefs, and if a net is passed over a reef it is quite likely it will become entangled in projecting coral. On the other hand, if a basket is dropped as a trap, it does not suffer as much damage from the coral„

Between the shore and the reef adjacent to the

Palmetto Grove site is a stretch of clear water with a smooth sandy bottom. I have not noticed many fishes in this area, although some small ones were observed in back of the beach breakers. An occasional barracuda will swim in 97 there also. Although this clear-uiater area would have been an excellent place to use nets, there can be no doubt

t that the place to get fish is on the reef Itself. Because most of the reef just off shore from the Palmetto Grove site is made up of algae, there is a very high percentage of algae-eating fish there. They may be speared or caught with traps made of baskets.

In summary, the shells were used either as net or basketry-trap weights for catching fish. There is no apparent pattern to making the weights, except for the perforations for suspension. No other kinds of weights have been found (such as ceramic or stone weights as occur elsewhere in the Caribbean)„

Shaft Scrapers. Four cut-shell objects were found with notches in their edges (Fig. 15). They resemble the chert equivalent of spokeshaves or arrow-shaft smoothers, except in size. It is not certain whether or not the

Indians of San Salvador had the bow and arrow, although there is no reason to presume they did not. There are ethnographic data (discussed in Chapter l) on the use of darts. The notches measure about 11, 12, 12, and 18 mm. in diameter from one corner to the opposite, This means the object being scraped or smoothed would have to be smaller in diameter than the average pencil. As all the notch is not usually taken up with the shaft being scraped, then

I believe these particular items were used on something A

Figure 15. Shell Shaft Scrapers. the size of darts, not arrows„ Two of these shaft scrapers were found at the 20-30 cm. level, one at the

10-20 cm. level, and the fourth at the 0-10 cm. level.

One was in association with the limestone paint palette and grinding stone.

Strombus Shell Gouges (Fig. 16). There are four main types of shell gouges found at the Palmetto Grove site, none of which precisely duplicates the criteria for gouges established elsewhere in the Caribbean. These gouges are simply the whorl portion of the Strombus cigas shell with the apex and spire removed to expose the smooth strong columne11a and one or more whorls. In some cases a portion of the whorls has been removed leaving only a slightly concave-convex, triangular, scoop-like implement, pointed and sharp at one end, broad and tough at the other

Some of the gouges are only a few centimeters long. It is not clear whether they were hafted with the pointed end into a socket, but if they were, then the small gouges may actually be those that broke off in the hafting. Lengths for all kinds of gouges vary from 2.5 cm. to 15 cm., with no evidence of clustering.

It is not known precisely what these tools were used for, but they may have been used for removing the wood from the dugout canoes about which Columbus wrote.

They may also have been used for digging in the sand, or gouging out fragments of limestone, or for chipping off 100 chunks of coral. Both ends show signs of having been used.

The broad end shows more use marks than the pointed, per~ haps because of use as hoes or some other kind of digging implement. On about one-half of the specimens there is a slight grinding area on one side of the tip and in the same place in each case. This is not a facet, but is rough textured, as though the tip had been held at an angle on something wooden and then hammered. Such use could also have produced the use marks on the broad end of the shell. Gouge, Type A (Fig. 16 A, B, D, E). This

type most resembles our modern day chisel. It

is far longer than it is wide, averaging about

11 cm. in length by 4 cm, across the chisel end.

The sides are roughly parallel. The utensil is

sort of spoon shaped. Seventeen Type A gouges

have been found at Palmetto Grove.

Gouge, Type B (Fig. 16 C, F, G, I, J , P,

Q ). This kind is triangular in outline. Again

the broad end is battered and broken, with

occasional grinding surfaces and slight angular

mars appearing on the side of the tip. They range

in length from 4.2 cm. to 14.5 cm. The width at

the widest part is from 2 to 8 cm. Twenty-one

Type B gouges were found. ) OCX*' Figure 16. Strombus Shell Gouges. A , B, D ? E , Type A Gouge; C, F, G „ I» J » P , Q, Type B Gouge; L, 0, R , Type C Gouge; K, N, Type D Gouge. H , Type E Gouge„ 101

Figure 16 Strombus Shell Gouges 102 Table 9. Vertical Distribution of Shell Gouges and Tips„

Only 11 squares mere excavated to the 40-50 cm. depth. See Figure 16.

Excavation Levels in Cm. Shell Gouges — --- —------— — Total 0-10 10-20 20-30 30-40 40-50

Type A 5 10 1 1 17

Type B 4 12 3 2 21

Type C 1 9 25 7 1 43

Type D 1 2 12 1 1 17

Type E 1 8 1 1 10

Shell Tips ' 3 11 30 10 54 (Unclassified small fragments)

Total 5 32 97 23 5 162 Strombus Gouge, Type C (Fig. 16 L,

M, 0 s, R ) „ The chief difference between Type

B and Type C is that the columnar whorl is not cut away in the latter specimen. The broad end does not present a concave-convex spoon or scoop shape„ but several spiralling whorls. Again, the small ground area shows up on the side of the tip. Forty-three Type

C gouges were found. They range from 4-15 cm. long.

Strombus Gouge, Type D (Fig. 16 K , N).

The fourth type is similar to Types A and B , except that it does not have a pointed tip.

In the process of making one of the Types A ,

B, or C Gouges, the Indians first removed the top or spire from the conch, and then sliced the remainder vertically. Such a "slice" would bisect the thick, solid base of the columnella. The whorl that was actually used with Types A , B , or C , is probably the penultimate, and not so much the body whorl, because in the process of manufacture the outer whorl was removed. The penultimate whorl is thicker and tougher than the body whorl. On the other hand, the umbilicus remains on the Type D Strombus gouge, 104 Indicating that the body whorl must be the

whorl that was used. The use marks are

along the top of the concave-convex scoop.

No use marks appear at the bottom of the

whorl, however. Strombus Gouge, Type E (Fig. 16 H )„

Here the bottom of the whorl has been shaped

to form a sort of handle. The upper

portion is the same as Types A and B , and

in fact seems to show even greater evidence

. of wear.

Strombus Hammers. Strombus hammers (Fig. 17 D ,

E, F) were apparently the type held in the hand rather than by a haft. At least there are no hafting perfora­ tions. Two types of hammers were found, apex and columnella. In either case the surface of the shell had been battered and crushed in the area involved. It is important that in the case of the apex hammers a considerable degree of battering be observed before the shell is classified as a hammer. Conchs are frequently used to produce the small perforation that is used for getting at the muscle of the animal inside the shell. The spire of one shell is simply struck against the side of the living specimen and with a quick blow a suitable hole is punched through the test. The small holes in the top of conchs (Fig. 6, 14) could have been produced in that 105 manner. This blunts the spires of the conch that was used to make the perforation. If several conchs are opened in this manner, the spire soon becomes very rounded and appears very much like a conch hammer, In a few cases, both apex arid columnella exhibit use as hammers.

Charonia Vessels. Two Charonia sp. shells were cut in such a manner as to produce what appear to be cups or vessels (Fig. 17 B ). These are not the same as those found elsewhere in the Caribbean (Harrington 1921, Hoffman

1963) but simpler and perhaps less functional, In the

San Salvador case the back of the body whorl has, in effect, been sliced off. The cut is fairly clean. In one of the two examples the slice continues up to include the back of the penultimate whorl (Fig„ 17 B), It is entirely possible that these shells did not serve as cups, but as net or trap weights, although they are much lighter than the plentiful conch, and wouldn’t have been as practical.

Cassis Horn. One Cassis flammea horn (Fig. 18) was found in the 20-30 cm. level. In addition to the apex having been removed for blowing through the shell, the back of the body whorl has also been cut out, making it possible to produce two different notes,

Strombus Ladle. The body whorl of a Strombus oioas. cut in such a manner (Fig. 17 C) as to provide a cup and handle. No marked striations are to be seen under /oSa- Figure.17. Miscellaneous Shell Implements. A , Cassis chopper ? B, Charonia vessels C , Strombus ladle; D-F, Strombus hammers. 106

Figure 17. Miscellaneous Shell Implements. 107 Figure 18. Cassis Horn. 108 i the microscope, although a few light ones appear to cross the outside near the top of the shell. A similar shell object is pictured by Rouse (i960. Fig. 3 P) as being from fflanicuare, Venezuela, but he does not mention it in the text. In Cruxent and Rouse (1959, PI. 6, No. 8) is illustrated a portion of a "cup" from the Cubagua complex.

Cassis Chopper. All but the front, or face, of a

Cassis tuberosa shell was cut away. The flat portion remaining (Fig. 17 A) was apparently used as some form of axe or chopper. The lip was apparently the handle, the opposite side of the instrument being the business end.

The Cassis. or king helmet shell, is not as tough as

Strombus and I suspect either that this particular tool was used on soft goods or that its life was limited. The artifact was first classified as some sort of bowl or dish, perhaps for keeping ornaments and the like from being lost in the sand. This may have been,the case, but as J. Cameron

Greenleaf demonstrated to me it does "fit the hand" perfectly when held as an axe, and this would explain the chipping and breakage on the opposite edge.

Shell Beads. There are several kinds of shell beads. They are distinguished by the small hole in some portion of the shell, as opposed to shell tinklers, whose sides may be cut or broken away. Some precaution must be taken to be certain the hole was drilled through the shell by man and not by some animal. There are at least two ways 109 Table 10. Vertical Distribution of Shell Beads and Tinklers. •

Only 11 squares were excavated to 50 cm, See Figure 9. ,

Excavation Levels in Cm, Beads and Tinklers — —--- Total 0-10 10-20 20-30 30-40 40-50

Ollvella sp, 7 7 9 7 1 31

Oliva sp. Tinklers 2 2 2 1 7

TeUlna sp. Fragment 1 1 2

Chione cancellata 1 1

L impit 1 3 2 6

Oliva sp. Beads 1 1 2

Family Naticidae 2 3 5 (Moon Shell)

Calliostoma sp, 1 . 1

Cypraea sp. 1 1

Univalve (Unidentified) 1 1

Total 11 12 17 16 1 57 , 1X0 to do this: if the hole is irregular in outline (and a microscope may be needed to determine this), or if it is hour-glass-shape in profile. The animal that normally drills through shells to get at the meat leaves a very neat conical hole„ Because shells used for beads are so thin it is difficult to drill a perfectly round hole„

In some cases it is drilled most of the way and then punched out, or the shell is turned around and the hole is drilled from the opposite direction, the latter producing the hour-glass outline„

Shell beads were made from Oliva. Calliostoma.

Cypraea„ Chione„ Codakia. Telllna„ family Naticidae

(Moon shell), and two species of limpit (Fig, 9 A-I, L , M).

In addition to the above beads there are many

Qlivella shells whose tops have been broken off. There are often fragments of tiny round holes at the edge of the fracture, one hole on each side of the shell,

Qlivella shell beads are commonly found in both ethnographic and archaeological situations. Thirty-one Qlivella beads were found,

Oliva Tinklers, Shell tinklers are noise-making beads (Fig, 9 J, K , 0-R). A part of the shell is broken away and a groove sawed through one end until it produces a hole, When a number of the beads are strung together, they make tinkling sounds of varying notes. Seven tinklers or tinkler fragments were found, all Oliva species„ Ill .Food Shells. Throughout the excavation were found thousands of small shells of many different species„

It is assumed that these shells are the residue from meals and that they were either cooked collectively in stews or soups, or eaten individually„ Many of the shells are still in one piece, that is, unbroken, many others have one side broken away. It is further assumed that the unbroken shells were those used in stews and those with sides or tops broken away were the ones eaten individually.

Some of these shells were separated out and their distributions plotted. . ’

Upon completion of the excavation at Palmetto

Grove the area around the site was tested, but not once did we find shells in such numbers as were found in the occupation area. On the site several thousand shells may be excavated; outside the site only one or two were found.

Many species were probably used for stews. These includes Chione cancellata. C. pensylvanlca of all sizes,

Divaricella guadrisculcata. Cerion sp., Cepolis sp., Dona* denticulatus. limpits of various species? Cerlthlum sp., family Naticidae, Columbella sp., Barbatia cancellaria.

Tectarlus muricatus. Lima sp., Haminoea sp., and many kinds of chitons. It would have been an impossible task to survey the distribution of the snails that had been used individually for food, In this case, as opposed to stew . 112 snails, the shell is broken open. Throughout the site there were countless thousands of fragments of shells, from snails both big and little. One of the more common represented the Nerita sp., the nerites, and in particular

N. peloronta. Another common shell is the “dove-shell,"

Columbella sp., chiefly £. mercatoria.

A record was kept of the distribution of the latter species along with the apparent method of breaking into the shell in order to pluck out the meat (Fig„ 19).

A total of 551 dove shells were counted, which could be divided into ten types on the basis of the method of breaking into the shells. More shells were found in the

30-40 cm. level than the one above, perhaps because the shells are so small that they tend to shift downward in the soil, There were 147 at 30-40 cm. and 110 in the 20-30 cm. level. For the latter, the breakdown iss top, front (that is the area on the shell that has been broken away), 29 shells; top, 21; front, 38; front, bottom, 5; tiny hole just inside the aperture, 1; back, 4; bottom, 6; top back,

5? and top side, 1.

Miscellaneous, Finally there are many fragments of obviously cut and worn shell that should be mentioned, but which can not yet be isolated according to function.

Strombus apex-tips. This fragment consists of

the apex and spire of the Strombus and a portion

of the columnella. The base is similar to the 113 too 0 0 #0 Top, Front T op F ront #» 00 0# Front, Bottom Inside Back npczlA nor tU 11U r*1 qo 0$ 00# Bottom Top, Back ditiTop, Side

noSide

Figure 19. Dove Shell Artifacts.

This illustrates ten different methods of breaking into the common Dove shell in order to pluck out the meat. The black area indicates that portion of the shell that was removed. 114

shell gouge, but instead of a broad scoop

at the top, the apex is still attached„

These objects may have been used, perhaps

as picks, but more likely they are the raw

materials from which Strombus gouges were

made.

Strombus Celt Blanks. Several

fragments of the Strombus sp. shell were

found, each resembling the shape of a celt, i or celt-hammer. They are all crude, un­

smoothed , and apparently the blanks which

may have been made into celts. However,

no celts were found in the Palmetto Grove

site. It may be that a lack of a hard

enough grinding surface prevented the

Indians from finishing these objects to the

stage normally recognized as being a celt.

Thus, they may not be blanks, but San

Salvador versions of the finished celt.

Coral

Coral was used chiefly as an abrasive„ Two species were preferred, Acropora palmate and A, cervicornis.

Coral Hone (Fig. 10 G ). This is a large leaf of

A,, palmata or elk horn coral on which appear five deep grinding facets. It is 28.9 cm. long and 16 cm. wide at 1X5 the widest part. The most prominent grinding mark is 1 cm. deep and 3 cm. wide and is V-shaped with very slightly concave sides, The groove is about 14 cm. long. The next largest facet is roughly 1 cm. deep, approximately

2 cm. across the top and 12 cm. of it remains along a fractured edge. The three remaining grooves are small, being 'but a few millimeters wide and deep and between 60cm. and 90 cm. long. The entire surface of the coral has been abraded to the point where the little polyps have been completely ground off. The under surface has not been used and the polyps there are still sharp and prominent„

Coral Corn Shuckers (Fig„ 10 D ). Several A. cervicornis fragments are so thick as to deserve separate attention. They vary between 3.9-4.8 cm. in diameter, and probably come from the base of a growth of this species of coral. It is called a "corn shucker" because that is the present day use for these objects by the people living on

San Salvador. Without doubt they make ideal corn shuckers and may have been used as such by the aboriginal inhabi­ tants of Palmetto Grove, One specimen is 21 cm. long. The polyps have been ground off all around, particularly at the ends.

Coral Rasps (Fig. 10 A , C). Numerous fragments of

A_» palmata were isolated because the convex side had been ground smooth. That is, the polyps, normally raised 116 above the surface (Fig. 10 B ) had been ground off. They average about 1.5 cm. in thickness, but the usable surface varies considerably from rasp to rasp.

Coral Files (Fig. 10 E ), In addition, several pieces of A., cervicornis were counted because they too had been worn smooth. With this particular species the number may vary widely because of the unknown frequency of possible breakage. Some were worn more at one end than the other. They vary in length from 3.4 to 13.6 cm.

Problematical Coral Objects. One problematical coral object was found in the 20-30 cm. level. It is ovoid, flat, and well ground all over. It is not A. palmata. but one of the brain corals. It is 4.9 cm. wide by 7,8 cm. long, and about 2.3 cm. thick at the middle,

There are no scratches or striations on it, although all the polyps on all surfaces have been ground off. 11 is not water-worn as the septa within each polyp are still very sharp*and distinct.

Another object is essentially a much smaller version of the previous example. It is 2*4 cm. by 2.8 cm, by .9 cm. thick at the center. It is much more worn than the other one on all surfaces, although there are no scratches or striations on it, either. CHAPTER 5

SUMMARY AND CONCLUSIONS

More than 98 per cent of the 5,250 sherds

recovered were of a single ware, Palmetto Ware, which is

characterized by certain ceramic modes: red paste, soft texture, shell tempering, and thicknesb. Most of the

Palmetto Ware pottery was found at a depth between 20=30

cm, A H but a few of these sherds are plain; 12 are decorated by affixing strips to the sides or by pinching

the rims to form points or horns, A few specimens of

other types of pottery were found. These may be from

vessels traded into the site. The most numerous is a

reddish paste, grit-tempered ware.

In contrast to most sites excavated in the

Caribbean, Palmetto Grove seems to represent a "shell

culture," rather than a "ceramic culture," The several

kinds of dhell found are archaeologically abundant and

readily available today. They probably served as a sub­

stitute for hard rock which is scarce on the island. Shell

was used to produce tools such as gouges, celts, hammers,

shaft scrapers, and weights, as well as ornaments.

Archaeologists in the Caribbean area, indeed, in

most of the New World, have found it convenient, often 117 118 necessary, to emphasize ceramic remains„ Pottery has been our "ball and chain," so to speak. Like it or not uie can­ not do without it. Thus, when we are dealing with a culture of a sedentary nature, and one that is known not to be ore-ceramic. we find that pottery dominates the excavation, the classification, and finally the reports.

The entire chronology of the prehistory of the Caribbean is in some way based upon ceramics; even the non-pottery making cultures are usually assigned time periods on the basis of the absence of sherds in a site. Admittedly there .is a need to revise our methodology but it is not entirely the fault of past workers; the problem is that there is usually little else upon which interpretations can be made. The Palmetto Grove site points up the long over­ due necessity to expand our technological know-how about shell, stone, bone, and other non-ceramic remains. It is simply not a site that can be discussed only in terms of its pottery. Without doubt Palmetto Ware as a type is probably unique to the Bahamas, and perhaps the northern-central Bahamas at that. It is probably a highly

"one-track" ware. The limitations in our methodology are clear if we cannot shift from pottery to some other artifact even when that artifact is the dominant one found in the site. Opposition is anticipated to the classification pattern for the shell remains found at the 119 Palmetto Grove„ However, if the site had been known as pre-ceramic (no pottery at all, that is, Ciboney) it is my opinion that such a classification would have been completely acceptable.

The use of shell is far underrated in the

Caribbean, Shell, especially that of the abundant

Strombus qiqas is heavy, tough, and strong. Fragments of shell may not get to be massive, but can easily substitute for most stone tools. They can cut or work wood as quickly as can stone, and in many instances are far more durable. But the point is, when rocks of sufficient strength are not available and shell is, then apparently the culture will turn to the latter. Not only should we become more familiar with the technological problems con­ cerned with producing shell tools, but we could improve our knowledge of the cultures of the Caribbean by learning the stylistic variations therein.

Hard stone artifacts at Palmetto Grove are rare, but some stone beads, a chopper or planer, and a number of barrel-shaped stone artifacts were found. The latter may have been "blanks'1 for making beads, or perhaps parts of some sort of game. Coral rasps and files were found, usually of the genus Acrooora. but they are not as numerous as in Lesser or Greater Antilles sites. A large Acropora hone was also recovered. 1 20

Of the faunal remains, 99 per cent are fishes,

and'of this over 50 per cent comprise algae-eating

fishes. most of the algae-eaters are parrotfishes.

One bone artifact, a pendant, was found. Thousands of

small sea shells and shells of land snails were found, apparently the remains of meals.

A somewhat different man-land relationship than

that normally found in the Caribbean was encountered at

the Palmetto Grove site. However it is viewed, the cul­

ture was extraordinarily close to the sea. The people

were very much dependent upon the sea for both food and

technology/.

It is difficult to determine what percentage of

food consumed came from the sea because the extent of

farming is unknown. The presence of griddles implies that

at least manioc was grown and made into bread or cakes,

and a kind of gruel. Eventually we should be able to

build up a file of ratios between land and sea that are

associated with different cultural groups. A complete, or

near complete, absence of land fauna remains indicates a

reliance upon fishes for meat at the Palmetto Grove,

moreover, a high percentage of algae-eating fish represented

in the remains closely correlates with the particular reef

immediately adjacent to the site and indicates that the

occupants of the site did not have to travel very far for

their protein. Other reefs around the island are not so 121 heavily populated with algae colonies, and there the omnivorous and carnivorous fishes are relatively more common.

The value of a diet chiefly composed of algae- eating fishes, as opposed to omnivorous fishes, would be a good subject for study. Because of the reduction in the number of steps between the production and consumption of energy, it would seem that the parrotfishes would be more beneficial than the omnivorous types.

Technologically, people of the Palmetto Grove site utilized the hard shells of various for tools and ornaments. Because impressions of basketry are fpund on the sides of some potsherds, we know they used local reeds and palm fronds to produce baskets and matting, Uie also know from ethnographic accounts that Arawakan-speaking peoples on the whole are reputed for their fine basketry.

The local soil known as Pineapple Loam was probably used for production of pottery, and may have been used for growing of crops.

In summary, it would seem that the adaptation of the people at the Palmetto Grove site was probably very close to that of the people living there today. Food was probably produced by a combination of farming and fishing

(including shell-fishing); villages werd probably amorphous, being scattered houses near the seashore. The day-to-day 1 2 2 existence was in all likelihood unhariied by problems of lack of shelter or something to eat. Shelter was apparently in the form of huts made of local trees and palmetto branches.

Conclusions

The people who deposited the remains at the

Palmetto Grove on San Salvador may have done so between

A.D. 850 and A.D. 1200 (Table ll). Thus, they probably never saw the three Spanish ships heralding the discovery of America by the Old World, nor did they experience the extinction of their identity as a cultural sub-group or breeding population.

One buff sherd of a sandy-paste ware, with incised sigmoid design was found at Palmetto Grove. This par­ ticular element of decoration is unusual except in sites of the Ostionoid level of time in the Greater Antilles.

The element is found in the Virgin Islands, Puerto Rico, and Jamaica. It is a mode that is part of a ceramic horizon normally considered ancestral to the Meillac style which developed in Hispaniola and which has been found in the Bahamas. The significance lies in the possibility of a pre-Meillac movement into the Bahamas, a movement much earlier than has heretofore been suspected. If true, the movement must have been after A.D. 7-800 when Ostiones Table 11. Suggested Chronology for Prehistoric Caribbean Area

Northern Turks a Dominlcon PUERTO RICO Virgin Haiti St. LuciaAntigua Jamaica Republic Islands Bahamas Caicos Eastern

■ x ■ ESPER FANNIS ■ ' m CARRIER eocA CAPA ANZA 1300 WHITE// -1300 MARL/> 1200 CHOC -1100 % SANTA 1000 ELENA 1000 LITTL p a l m e t t z .MEILLAC MASS­ MILL -900 9 0 0 - * rTv ER ACRE

MACADY NADEL OSTIONES TROU

6 00 LATE -600 CORAL CUEVAS BAY

5 0 0- -500

40 0 -4 00 EARLY

3 0 0- HACIENDE GRANDE CORAL -300 BAY 124 began in the Greater Antilles (the Virgin Islands?), and before A.D„ 1200 when the Meillac style developed.

Rouse believes that the Meillac developed out of an Ostionoid base. Styles coeval to late Ostiones are

Santa Elena in eastern Puerto Rico and Magens in the

Virgin Islands. There is some agreement that this is a broad ceramic horizon, found on many of the islands and which is represented by a number of specific attributes.

As far as we are concerned here these attributes include: rim points or horns, strip lugs, lunar, wedge-shaped lugs, and incising across the vessel lip. All of these traits are to be found on Palmetto Ware or associated ceramics

(in addition to the sigmoid applique'' design). Thus it would seem that Palmetto Ware may be a part of the

Ostiones-Santa Elena-fflagehs time horizon. This time horizon is dated (by radiocarbon) between A„D. 900 and 1200.

Ostiones itself, the type site for this series of ceramic complexes, begins earlier (A.D. 7-800), but we are con­ cerned here with the late development of the series which does not begin until around A.D. 900.

Even if the Palmetto Grove site was occupied at a later time, it certainly was not as late as 12 October

1492. There are no trade objects at all from the occupa­ tion level, and none have been reported from any previous excavation. Some modern green glass and metal fragments 125 occur in the first few centimeters but these are apparently quite recent. The relationship between the Bahamas and the

Greater Antilles during the Ostionoid Horizon may extend into the Lesser Antillean islands. After recent excavations on Antigua and St. Kitts in the Leeward Islands, I was struck by the close similarity between sherds from the Mill

Reef Complex on Antigua and the Botany pottery from the

Virgin Islands. The similarity was more in the paste and surface finish than in decoration. It may be significant that there is some similarity between the Virgin Islands

Botany sherds and the Palmetto Ware from San Salvador, although the resemblance here is in the decoration, not • the paste. Indeed, no paste in all the West Indies is apparently like the crude San Salvador Palmetto Ware.

It may be that the cultural manifestations we are finding at the time of late Ostiones is pan-Antillean and represents a movement or reshuffling of peoples within or into the West Indies. At least it would seem that people had moved into the Bahamas some time before the subsequent

Meillac styling developed out of the Ostiones.

In summary, it seems that the Palmetto Grove site is part of a much wider cultural group, evidence for which we find in both the Greater and Lesser Antilles. This group, or at least the collective evidence for this group, 126 has been named the Ostlones Series. It probably dates between A.D. 750 and 1200 with a late phase beginning around A.D. 900. REFERENCES

Bahamas Chamber of Commerce

1952 Announcement„ Nassau.

Barton, G, T„

1953 The Prehistory of the Barbados. The Advocate

Co., Ltd. Barbados.

Brooks, UJ. K.

1688 On the Lucayan Indians. Memoirs of the

National Academy of Sciences. Vol. 4, Part 2,

pp. 215-23. Washington.

Bullen, Ripley P.

1963 Ceramic Periods of St. Thomas and St. John

Islands, Virgin Islands. American Studies.

No. 4. William L . Bryant Foundation, Orlando.

1964 The Archaeology of Grenada, West Indies.

Contributions of the Florida State Museum.

Social Sciences. No, 11. Gainesville.

Cruxent, J „ M . and Irving Rouse

1959 An Archeological Chronology of Venezuela.

Social Science Monographs. No. 6, Vol. 2.

Pan American Union, Washington. 127 128 DeBooy„ Thsodoor 1912 Lucayan Remains on the Caicos Islands,

American Anthropologist. Vol. 14, No. 1,

pp. 81-105. Lancaster,

1913 Lucayan Artifacts From the Bahamas.

American Anthropologist. Vol. 15, No, 1,

pp. 1-7. Lancaster.

Goggin, J . M . 1939 An Anthropological Reconnaissance of Andros

Island, Bahamas, American Antiquity. Vol. 5,

No. 1, pp. 21-6, Wenasha.

1960 The Spanish Olive Jar. Vale University

Publications in Anthropology. No, 62,

New Haven,

Gower, C. D,

1927 The Northern and Southern Affiliation of

Antillean Culture. Memoirs of the American

Anthropological Association. No. 35. Menasha,

Cranberry, Julian 1955 A Survey of Bahamian Archeology. MS,

Master*s Thesis, Department of Anthropology,

University of Florida, Gainesville.

1956 The Cultural Position of the Bahamas in 129 Caribbean Archaeology„ American Antiquity.

Vol. 22, No. 2, pp. 128-34. Salt Lake City.

1957 An Anthropological Reconnaissance of Bimini,

Bahamas. American Antiquity. Vol. 22, No. 4,

pp. 378-81. Salt Lake City.

Haag, William

1965 Pottery Typology in Certain Lesser Antilles.

American Antiquity. Vol. 31, No. 2, pp.

242-5. Salt Lake City.

Harrington, Mark R .

1921 Cuba Before Columbus. Indian Notes and

Monographs. Part 1, Vols. 1-2, Museum of the

American Indian, Heye Foundation, Miscellaneous

Series. New York.

Hatt, Gudmund

1924 Archaeology of the Virgin Islands.

Proceedings of the Twenty-First International

Congress of Americanists. First Part, pp. 29-

42. The Hague.

Hoffman, Charles A., Jr.

1963 Archeological Investigations on Antigua,

West Indies. MS, Master's Thesis, Department

of Anthropology, University of Florida,

Gainesville. 130 Holmes, William H„

1894 Caribbean Influence in the Prehistoric Art

of Southern States„ American Anthropologist,

o. s., Vol. 7, No,; 1, pp. 71-9. Washington.

Krieger, Herbert W.

1937 The Bahama Islands and Their Prehistoric

Population. Explorations and Field Work of

the Smithsonian Instltutlon in 1936. pp. 93-8.

Washington.

Las Casas, Bartolomeide

1952 The Tears of the Indians. Translated by

John Phillips. Academic Reprints (Repro­

duction of an original English edition of

1656)„ Stanford.

Loven, Sven

1935 Origins of the Tainan Culture. West Indies.

Elanders Boktrycheri Aktiebolag. Goteborg.

Major, R . H , Editor and Translator

1870 Select Letters of Christopher Columbus With

Other Original Documents. Works issued by

the Hakluyt Society, No. 43, London. 131

Mason, Otis T .

1904 Aboriginal American Basketry, U.S, National

Museum Report. 1902. Washington.

McKusick, Marshall

1960 The Distribution of Ceramic Styles in the

Lesser Antilles. MS, Doctoral Dissertation, .

Department of Anthropology, Yale University,

New Haven,

Morison, Samuel E.liot^

1942 Admiral of the Ocean Sea (Abridged), Little,

Brown and Co., Boston.

Osgood, Cornelius

1942 The Cibdney Culture of Cayo Redondo, Cuba.

Yale University Publications in Anthropology.

No, 25. New Haven.

Rainey, Froelich G.

1934 Diary Beginning January 22, 1934, Upon Arrival

in Port au Prince, Haiti. MS, Peabody Museum,

Yale University. New Haven.

1940 Porto Rican Archeology. Scientific Survey of

Porto Rico and the Virgin Islands. Vol. 18,

Part 1. New York Academy of Sciences,

New York. 132

Rainey, F . G „

1941 Excavations in the Ft. Liberte Region,

Haiti. Yale University Publications in

AnthropolQQv. No. 23. New Haven.

Rouse, Irving 1939 Prehistory in Haiti $ A Study in Method.

Yale University Publications in Anthropology.

No. 21. New Haven„

1941 Culture of the Ft. Liberte Region, Haiti.

Yale University Publications in Anthropology.

No. 24. New Haven.

1948 The uJest Indies. In "Handbook of South

American Indians," edited by J„ H . Steward.

Bureau of American Ethnology. Bulletin 143,

Vol. 4, Part 3, pp. 495-565. Washington.

1949 The Southeast and the West Indies. In The

Florida Indian and His Neighbors. Edited by

J. W. Griffin, pp. 117-37, Winter Park.

1952 Porto Rican Prehistory. Scientific Survey

of Porto Rico and the Virgin Islands. Vol. 18,

Parts 3-4, New York Academy of Sciences,

New York.

I960 The Entry of Man Into the West Indies.

Yale University Publications in Anthropology.

No, 61. New Haven. 133

Rouse, Irving 1962 The Intermediate Area, Amazonia and the

Caribbean Area. In "Courses Toward Urban

Life," edited by R. L . Braidwood and G, R.

Willey, pp. 34-59. Viking Fund Publications

in Anthropology. No. 32, Wenner-Gren

Foundation for Anthropological Research,

New York.

1963 Dating of Caribbean Cultures. MS, on file.

Department of Anthropology, Yale University,

New Haven.

1964 Prehistory of the West Indies. Science.

Vol. 144, No. 3618, pp. 499-513. Washington.

Shattuck, George Burbank

1905 The Bahama Islands. Geographical Society x of Baltimore. MacMillan, New York.

Smithsonian Institution

1938 Annual Report. . .for the Year Ending June

30. 1957. Washington.

1948 Annual Report. . .for the Year Ending June

30. 1947. Washington. I Yearbook for Agriculture

1941 Climate and Man. United States Department

of Agriculture. Washington,